47 import xml
.etree
.ElementTree
50 from . import traversal
52 from ..compat
import functools
# isort: split
53 from ..compat
import (
54 compat_etree_fromstring
,
56 compat_HTMLParseError
,
60 from ..dependencies
import brotli
, certifi
, websockets
, xattr
61 from ..socks
import ProxyType
, sockssocket
63 # This is not clearly defined otherwise
64 compiled_regex_type
= type(re
.compile(''))
67 def random_user_agent():
68 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
109 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
112 SUPPORTED_ENCODINGS
= [
116 SUPPORTED_ENCODINGS
.append('br')
119 'User-Agent': random_user_agent(),
120 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
121 'Accept-Language': 'en-us,en;q=0.5',
122 'Sec-Fetch-Mode': 'navigate',
127 'Safari': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27',
131 NO_DEFAULT
= object()
132 IDENTITY
= lambda x
: x
134 ENGLISH_MONTH_NAMES
= [
135 'January', 'February', 'March', 'April', 'May', 'June',
136 'July', 'August', 'September', 'October', 'November', 'December']
139 'en': ENGLISH_MONTH_NAMES
,
141 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
142 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
143 # these follow the genitive grammatical case (dopełniacz)
144 # some websites might be using nominative, which will require another month list
145 # https://en.wikibooks.org/wiki/Polish/Noun_cases
146 'pl': ['stycznia', 'lutego', 'marca', 'kwietnia', 'maja', 'czerwca',
147 'lipca', 'sierpnia', 'września', 'października', 'listopada', 'grudnia'],
150 # From https://github.com/python/cpython/blob/3.11/Lib/email/_parseaddr.py#L36-L42
152 'UT': 0, 'UTC': 0, 'GMT': 0, 'Z': 0,
153 'AST': -4, 'ADT': -3, # Atlantic (used in Canada)
154 'EST': -5, 'EDT': -4, # Eastern
155 'CST': -6, 'CDT': -5, # Central
156 'MST': -7, 'MDT': -6, # Mountain
157 'PST': -8, 'PDT': -7 # Pacific
160 # needed for sanitizing filenames in restricted mode
161 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
162 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
163 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
193 '%Y-%m-%d %H:%M:%S.%f',
194 '%Y-%m-%d %H:%M:%S:%f',
197 '%Y-%m-%dT%H:%M:%SZ',
198 '%Y-%m-%dT%H:%M:%S.%fZ',
199 '%Y-%m-%dT%H:%M:%S.%f0Z',
201 '%Y-%m-%dT%H:%M:%S.%f',
204 '%b %d %Y at %H:%M:%S',
206 '%B %d %Y at %H:%M:%S',
210 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
211 DATE_FORMATS_DAY_FIRST
.extend([
221 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
222 DATE_FORMATS_MONTH_FIRST
.extend([
230 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
231 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>\s
*(?P
<json_ld
>{.+?}|\
[.+?\
])\s
*</script
>'
233 NUMBER_RE = r'\d
+(?
:\
.\d
+)?
'
237 def preferredencoding():
238 """Get preferred encoding.
240 Returns the best encoding scheme for the system, based on
241 locale.getpreferredencoding() and some further tweaks.
244 pref = locale.getpreferredencoding()
252 def write_json_file(obj, fn):
253 """ Encode obj as JSON and write it to fn, atomically if possible """
255 tf = tempfile.NamedTemporaryFile(
256 prefix=f'{os.path.basename(fn)}
.', dir=os.path.dirname(fn),
257 suffix='.tmp
', delete=False, mode='w
', encoding='utf
-8')
261 json.dump(obj, tf, ensure_ascii=False)
262 if sys.platform == 'win32
':
263 # Need to remove existing file on Windows, else os.rename raises
264 # WindowsError or FileExistsError.
265 with contextlib.suppress(OSError):
267 with contextlib.suppress(OSError):
270 os.chmod(tf.name, 0o666 & ~mask)
271 os.rename(tf.name, fn)
273 with contextlib.suppress(OSError):
278 def find_xpath_attr(node, xpath, key, val=None):
279 """ Find the xpath xpath[@key=val] """
280 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
281 expr = xpath + ('[@%s]' % key if val is None else f"[@{key}='{val}
']")
282 return node.find(expr)
284 # On python2.6 the xml.etree.ElementTree.Element methods don't support
285 # the namespace parameter
288 def xpath_with_ns(path
, ns_map
):
289 components
= [c
.split(':') for c
in path
.split('/')]
293 replaced
.append(c
[0])
296 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
297 return '/'.join(replaced
)
300 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
301 def _find_xpath(xpath
):
302 return node
.find(xpath
)
304 if isinstance(xpath
, str):
305 n
= _find_xpath(xpath
)
313 if default
is not NO_DEFAULT
:
316 name
= xpath
if name
is None else name
317 raise ExtractorError('Could not find XML element %s' % name
)
323 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
324 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
325 if n
is None or n
== default
:
328 if default
is not NO_DEFAULT
:
331 name
= xpath
if name
is None else name
332 raise ExtractorError('Could not find XML element\'s text %s' % name
)
338 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
339 n
= find_xpath_attr(node
, xpath
, key
)
341 if default
is not NO_DEFAULT
:
344 name
= f
'{xpath}[@{key}]' if name
is None else name
345 raise ExtractorError('Could not find XML attribute %s' % name
)
351 def get_element_by_id(id, html
, **kwargs
):
352 """Return the content of the tag with the specified ID in the passed HTML document"""
353 return get_element_by_attribute('id', id, html
, **kwargs
)
356 def get_element_html_by_id(id, html
, **kwargs
):
357 """Return the html of the tag with the specified ID in the passed HTML document"""
358 return get_element_html_by_attribute('id', id, html
, **kwargs
)
361 def get_element_by_class(class_name
, html
):
362 """Return the content of the first tag with the specified class in the passed HTML document"""
363 retval
= get_elements_by_class(class_name
, html
)
364 return retval
[0] if retval
else None
367 def get_element_html_by_class(class_name
, html
):
368 """Return the html of the first tag with the specified class in the passed HTML document"""
369 retval
= get_elements_html_by_class(class_name
, html
)
370 return retval
[0] if retval
else None
373 def get_element_by_attribute(attribute
, value
, html
, **kwargs
):
374 retval
= get_elements_by_attribute(attribute
, value
, html
, **kwargs
)
375 return retval
[0] if retval
else None
378 def get_element_html_by_attribute(attribute
, value
, html
, **kargs
):
379 retval
= get_elements_html_by_attribute(attribute
, value
, html
, **kargs
)
380 return retval
[0] if retval
else None
383 def get_elements_by_class(class_name
, html
, **kargs
):
384 """Return the content of all tags with the specified class in the passed HTML document as a list"""
385 return get_elements_by_attribute(
386 'class', r
'[^\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
387 html, escape_value=False)
390 def get_elements_html_by_class(class_name, html):
391 """Return the html of all tags with the specified class in the passed HTML document as a list"""
392 return get_elements_html_by_attribute(
393 'class', r'[^
\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
394 html, escape_value=False)
397 def get_elements_by_attribute(*args, **kwargs):
398 """Return the content of the tag with the specified attribute in the passed HTML document"""
399 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
402 def get_elements_html_by_attribute(*args, **kwargs):
403 """Return the html of the tag with the specified attribute in the passed HTML document"""
404 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
407 def get_elements_text_and_html_by_attribute(attribute, value, html, *, tag=r'[\w
:.-]+', escape_value=True):
409 Return the text (content) and the html (whole) of the tag with the specified
410 attribute in the passed HTML document
415 quote = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
417 value = re.escape(value) if escape_value else value
419 partial_element_re = rf'''(?x
)
421 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
422 \s{re.escape(attribute)}\s*=\s*(?P<_q>['"]{quote})(?-x:{value})(?P=_q)
425 for m in re.finditer(partial_element_re, html):
426 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
429 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
434 class HTMLBreakOnClosingTagParser(html.parser.HTMLParser):
436 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
437 closing tag for the first opening tag it has encountered, and can be used
441 class HTMLBreakOnClosingTagException(Exception):
445 self.tagstack = collections.deque()
446 html.parser.HTMLParser.__init__(self)
451 def __exit__(self, *_):
455 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
456 # so data remains buffered; we no longer have any interest in it, thus
457 # override this method to discard it
460 def handle_starttag(self, tag, _):
461 self.tagstack.append(tag)
463 def handle_endtag(self, tag):
464 if not self.tagstack:
465 raise compat_HTMLParseError('no tags
in the stack
')
467 inner_tag = self.tagstack.pop()
471 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
472 if not self.tagstack:
473 raise self.HTMLBreakOnClosingTagException()
476 # XXX: This should be far less strict
477 def get_element_text_and_html_by_tag(tag, html):
479 For the first element with the specified tag in the passed HTML document
480 return its' content (text
) and the whole
element (html
)
482 def find_or_raise(haystack, needle, exc):
484 return haystack.index(needle)
487 closing_tag = f'</{tag}>'
488 whole_start = find_or_raise(
489 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
490 content_start = find_or_raise(
491 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
492 content_start += whole_start + 1
493 with HTMLBreakOnClosingTagParser() as parser:
494 parser.feed(html[whole_start:content_start])
495 if not parser.tagstack or parser.tagstack[0] != tag:
496 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
497 offset = content_start
498 while offset < len(html):
499 next_closing_tag_start = find_or_raise(
500 html[offset:], closing_tag,
501 compat_HTMLParseError(f'closing {tag} tag not found'))
502 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
504 parser.feed(html[offset:offset + next_closing_tag_end])
505 offset += next_closing_tag_end
506 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
507 return html[content_start:offset + next_closing_tag_start], \
508 html[whole_start:offset + next_closing_tag_end]
509 raise compat_HTMLParseError('unexpected end of html')
512 class HTMLAttributeParser(html.parser.HTMLParser):
513 """Trivial HTML parser to gather the attributes
for a single element
"""
517 html.parser.HTMLParser.__init__(self)
519 def handle_starttag(self, tag, attrs):
520 self.attrs = dict(attrs)
521 raise compat_HTMLParseError('done')
524 class HTMLListAttrsParser(html.parser.HTMLParser):
525 """HTML parser to gather the attributes
for the elements of a
list"""
528 html.parser.HTMLParser.__init__(self)
532 def handle_starttag(self, tag, attrs):
533 if tag == 'li' and self._level == 0:
534 self.items.append(dict(attrs))
537 def handle_endtag(self, tag):
541 def extract_attributes(html_element):
542 """Given a string
for an HTML element such
as
544 a
="foo" B
="bar" c
="&98;az" d
=boz
545 empty
= noval entity
="&"
548 Decode
and return a dictionary of attributes
.
550 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
551 'empty': '', 'noval': None, 'entity': '&',
552 'sq': '"', 'dq': '\''
555 parser = HTMLAttributeParser()
556 with contextlib.suppress(compat_HTMLParseError):
557 parser.feed(html_element)
562 def parse_list(webpage):
563 """Given a string
for an series of HTML
<li
> elements
,
564 return a dictionary of their attributes
"""
565 parser = HTMLListAttrsParser()
571 def clean_html(html):
572 """Clean an HTML snippet into a readable string
"""
574 if html is None: # Convenience for sanitizing descriptions etc.
577 html = re.sub(r'\s+', ' ', html)
578 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
579 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
581 html = re.sub('<.*?>', '', html)
582 # Replace html entities
583 html = unescapeHTML(html)
587 class LenientJSONDecoder(json.JSONDecoder):
589 def __init__(self, *args, transform_source=None, ignore_extra=False, close_objects=0, **kwargs):
590 self.transform_source, self.ignore_extra = transform_source, ignore_extra
591 self._close_attempts = 2 * close_objects
592 super().__init__(*args, **kwargs)
595 def _close_object(err):
596 doc = err.doc[:err.pos]
597 # We need to add comma first to get the correct error message
598 if err.msg.startswith('Expecting \',\''):
600 elif not doc.endswith(','):
603 if err.msg.startswith('Expecting property name'):
604 return doc[:-1] + '}'
605 elif err.msg.startswith('Expecting value'):
606 return doc[:-1] + ']'
609 if self.transform_source:
610 s = self.transform_source(s)
611 for attempt in range(self._close_attempts + 1):
613 if self.ignore_extra:
614 return self.raw_decode(s.lstrip())[0]
615 return super().decode(s)
616 except json.JSONDecodeError as e:
619 elif attempt < self._close_attempts:
620 s = self._close_object(e)
623 raise type(e)(f'{e.msg} in {s[e.pos-10:e.pos+10]!r}', s, e.pos)
624 assert False, 'Too many attempts to decode JSON'
627 def sanitize_open(filename, open_mode):
628 """Try to
open the given filename
, and slightly tweak it
if this fails
.
630 Attempts to
open the given filename
. If this fails
, it tries to change
631 the filename slightly
, step by step
, until it
's either able to open it
632 or it fails and raises a final exception, like the standard open()
635 It returns the tuple (stream, definitive_file_name).
638 if sys.platform == 'win32
':
641 # stdout may be any IO stream, e.g. when using contextlib.redirect_stdout
642 with contextlib.suppress(io.UnsupportedOperation):
643 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
644 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
646 for attempt in range(2):
649 if sys.platform == 'win32
':
650 # FIXME: An exclusive lock also locks the file from being read.
651 # Since windows locks are mandatory, don't lock the
file on
windows (for now
).
652 # Ref: https://github.com/yt-dlp/yt-dlp/issues/3124
653 raise LockingUnsupportedError()
654 stream
= locked_file(filename
, open_mode
, block
=False).__enter
__()
656 stream
= open(filename
, open_mode
)
657 return stream
, filename
658 except OSError as err
:
659 if attempt
or err
.errno
in (errno
.EACCES
,):
661 old_filename
, filename
= filename
, sanitize_path(filename
)
662 if old_filename
== filename
:
666 def timeconvert(timestr
):
667 """Convert RFC 2822 defined time string into system timestamp"""
669 timetuple
= email
.utils
.parsedate_tz(timestr
)
670 if timetuple
is not None:
671 timestamp
= email
.utils
.mktime_tz(timetuple
)
675 def sanitize_filename(s
, restricted
=False, is_id
=NO_DEFAULT
):
676 """Sanitizes a string so it could be used as part of a filename.
677 @param restricted Use a stricter subset of allowed characters
678 @param is_id Whether this is an ID that should be kept unchanged if possible.
679 If unset, yt-dlp's new sanitization rules are in effect
684 def replace_insane(char
):
685 if restricted
and char
in ACCENT_CHARS
:
686 return ACCENT_CHARS
[char
]
687 elif not restricted
and char
== '\n':
689 elif is_id
is NO_DEFAULT
and not restricted
and char
in '"*:<>?|/\\':
690 # Replace with their full-width unicode counterparts
691 return {'/': '\u29F8', '\\': '\u29f9'}
.get(char
, chr(ord(char
) + 0xfee0))
692 elif char
== '?' or ord(char
) < 32 or ord(char
) == 127:
695 return '' if restricted
else '\''
697 return '\0_\0-' if restricted
else '\0 \0-'
698 elif char
in '\\/|*<>':
700 if restricted
and (char
in '!&\'()[]{}$;`^,#' or char
.isspace() or ord(char
) > 127):
704 # Replace look-alike Unicode glyphs
705 if restricted
and (is_id
is NO_DEFAULT
or not is_id
):
706 s
= unicodedata
.normalize('NFKC', s
)
707 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
) # Handle timestamps
708 result
= ''.join(map(replace_insane
, s
))
709 if is_id
is NO_DEFAULT
:
710 result
= re
.sub(r
'(\0.)(?:(?=\1)..)+', r
'\1', result
) # Remove repeated substitute chars
711 STRIP_RE
= r
'(?:\0.|[ _-])*'
712 result
= re
.sub(f
'^\0.{STRIP_RE}|{STRIP_RE}\0.$', '', result
) # Remove substitute chars from start/end
713 result
= result
.replace('\0', '') or '_'
716 while '__' in result
:
717 result
= result
.replace('__', '_')
718 result
= result
.strip('_')
719 # Common case of "Foreign band name - English song title"
720 if restricted
and result
.startswith('-_'):
722 if result
.startswith('-'):
723 result
= '_' + result
[len('-'):]
724 result
= result
.lstrip('.')
730 def sanitize_path(s
, force
=False):
731 """Sanitizes and normalizes path on Windows"""
732 if sys
.platform
== 'win32':
734 drive_or_unc
, _
= os
.path
.splitdrive(s
)
740 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
744 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
745 for path_part
in norm_path
]
747 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
748 elif force
and s
and s
[0] == os
.path
.sep
:
749 sanitized_path
.insert(0, os
.path
.sep
)
750 return os
.path
.join(*sanitized_path
)
753 def sanitize_url(url
, *, scheme
='http'):
754 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
755 # the number of unwanted failures due to missing protocol
758 elif url
.startswith('//'):
759 return f
'{scheme}:{url}'
760 # Fix some common typos seen so far
762 # https://github.com/ytdl-org/youtube-dl/issues/15649
763 (r
'^httpss://', r
'https://'),
764 # https://bx1.be/lives/direct-tv/
765 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
767 for mistake
, fixup
in COMMON_TYPOS
:
768 if re
.match(mistake
, url
):
769 return re
.sub(mistake
, fixup
, url
)
773 def extract_basic_auth(url
):
774 parts
= urllib
.parse
.urlsplit(url
)
775 if parts
.username
is None:
777 url
= urllib
.parse
.urlunsplit(parts
._replace
(netloc
=(
778 parts
.hostname
if parts
.port
is None
779 else '%s:%d' % (parts
.hostname
, parts
.port
))))
780 auth_payload
= base64
.b64encode(
781 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode())
782 return url
, f
'Basic {auth_payload.decode()}'
785 def sanitized_Request(url
, *args
, **kwargs
):
786 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
787 if auth_header
is not None:
788 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
789 headers
['Authorization'] = auth_header
790 return urllib
.request
.Request(url
, *args
, **kwargs
)
794 """Expand shell variables and ~"""
795 return os
.path
.expandvars(compat_expanduser(s
))
798 def orderedSet(iterable
, *, lazy
=False):
799 """Remove all duplicates from the input iterable"""
801 seen
= [] # Do not use set since the items can be unhashable
807 return _iter() if lazy
else list(_iter())
810 def _htmlentity_transform(entity_with_semicolon
):
811 """Transforms an HTML entity to a character."""
812 entity
= entity_with_semicolon
[:-1]
814 # Known non-numeric HTML entity
815 if entity
in html
.entities
.name2codepoint
:
816 return chr(html
.entities
.name2codepoint
[entity
])
818 # TODO: HTML5 allows entities without a semicolon.
819 # E.g. 'Éric' should be decoded as 'Éric'.
820 if entity_with_semicolon
in html
.entities
.html5
:
821 return html
.entities
.html5
[entity_with_semicolon
]
823 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
825 numstr
= mobj
.group(1)
826 if numstr
.startswith('x'):
828 numstr
= '0%s' % numstr
831 # See https://github.com/ytdl-org/youtube-dl/issues/7518
832 with contextlib
.suppress(ValueError):
833 return chr(int(numstr
, base
))
835 # Unknown entity in name, return its literal representation
836 return '&%s;' % entity
842 assert isinstance(s
, str)
845 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
848 def escapeHTML(text
):
851 .replace('&', '&')
852 .replace('<', '<')
853 .replace('>', '>')
854 .replace('"', '"')
855 .replace("'", ''')
859 def process_communicate_or_kill(p
, *args
, **kwargs
):
860 deprecation_warning(f
'"{__name__}.process_communicate_or_kill" is deprecated and may be removed '
861 f
'in a future version. Use "{__name__}.Popen.communicate_or_kill" instead')
862 return Popen
.communicate_or_kill(p
, *args
, **kwargs
)
865 class Popen(subprocess
.Popen
):
866 if sys
.platform
== 'win32':
867 _startupinfo
= subprocess
.STARTUPINFO()
868 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
873 def _fix_pyinstaller_ld_path(env
):
874 """Restore LD_LIBRARY_PATH when using PyInstaller
875 Ref: https://github.com/pyinstaller/pyinstaller/blob/develop/doc/runtime-information.rst#ld_library_path--libpath-considerations
876 https://github.com/yt-dlp/yt-dlp/issues/4573
878 if not hasattr(sys
, '_MEIPASS'):
882 orig
= env
.get(f
'{key}_ORIG')
888 _fix('LD_LIBRARY_PATH') # Linux
889 _fix('DYLD_LIBRARY_PATH') # macOS
891 def __init__(self
, *args
, env
=None, text
=False, **kwargs
):
893 env
= os
.environ
.copy()
894 self
._fix
_pyinstaller
_ld
_path
(env
)
896 self
.__text
_mode
= kwargs
.get('encoding') or kwargs
.get('errors') or text
or kwargs
.get('universal_newlines')
898 kwargs
['universal_newlines'] = True # For 3.6 compatibility
899 kwargs
.setdefault('encoding', 'utf-8')
900 kwargs
.setdefault('errors', 'replace')
901 super().__init
__(*args
, env
=env
, **kwargs
, startupinfo
=self
._startupinfo
)
903 def communicate_or_kill(self
, *args
, **kwargs
):
905 return self
.communicate(*args
, **kwargs
)
906 except BaseException
: # Including KeyboardInterrupt
907 self
.kill(timeout
=None)
910 def kill(self
, *, timeout
=0):
913 self
.wait(timeout
=timeout
)
916 def run(cls
, *args
, timeout
=None, **kwargs
):
917 with cls(*args
, **kwargs
) as proc
:
918 default
= '' if proc
.__text
_mode
else b
''
919 stdout
, stderr
= proc
.communicate_or_kill(timeout
=timeout
)
920 return stdout
or default
, stderr
or default
, proc
.returncode
923 def encodeArgument(s
):
924 # Legacy code that uses byte strings
925 # Uncomment the following line after fixing all post processors
926 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, str, type(s))
927 return s
if isinstance(s
, str) else s
.decode('ascii')
930 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
933 def timetuple_from_msec(msec
):
934 secs
, msec
= divmod(msec
, 1000)
935 mins
, secs
= divmod(secs
, 60)
936 hrs
, mins
= divmod(mins
, 60)
937 return _timetuple(hrs
, mins
, secs
, msec
)
940 def formatSeconds(secs
, delim
=':', msec
=False):
941 time
= timetuple_from_msec(secs
* 1000)
943 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
945 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
947 ret
= '%d' % time
.seconds
948 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
951 def _ssl_load_windows_store_certs(ssl_context
, storename
):
952 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
954 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
955 if encoding
== 'x509_asn' and (
956 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
957 except PermissionError
:
960 with contextlib
.suppress(ssl
.SSLError
):
961 ssl_context
.load_verify_locations(cadata
=cert
)
964 def make_HTTPS_handler(params
, **kwargs
):
965 opts_check_certificate
= not params
.get('nocheckcertificate')
966 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
967 context
.check_hostname
= opts_check_certificate
968 if params
.get('legacyserverconnect'):
969 context
.options |
= 4 # SSL_OP_LEGACY_SERVER_CONNECT
970 # Allow use of weaker ciphers in Python 3.10+. See https://bugs.python.org/issue43998
971 context
.set_ciphers('DEFAULT')
973 sys
.version_info
< (3, 10)
974 and ssl
.OPENSSL_VERSION_INFO
>= (1, 1, 1)
975 and not ssl
.OPENSSL_VERSION
.startswith('LibreSSL')
977 # Backport the default SSL ciphers and minimum TLS version settings from Python 3.10 [1].
978 # This is to ensure consistent behavior across Python versions, and help avoid fingerprinting
979 # in some situations [2][3].
980 # Python 3.10 only supports OpenSSL 1.1.1+ [4]. Because this change is likely
981 # untested on older versions, we only apply this to OpenSSL 1.1.1+ to be safe.
982 # LibreSSL is excluded until further investigation due to cipher support issues [5][6].
983 # 1. https://github.com/python/cpython/commit/e983252b516edb15d4338b0a47631b59ef1e2536
984 # 2. https://github.com/yt-dlp/yt-dlp/issues/4627
985 # 3. https://github.com/yt-dlp/yt-dlp/pull/5294
986 # 4. https://peps.python.org/pep-0644/
987 # 5. https://peps.python.org/pep-0644/#libressl-support
988 # 6. https://github.com/yt-dlp/yt-dlp/commit/5b9f253fa0aee996cf1ed30185d4b502e00609c4#commitcomment-89054368
989 context
.set_ciphers('@SECLEVEL=2:ECDH+AESGCM:ECDH+CHACHA20:ECDH+AES:DHE+AES:!aNULL:!eNULL:!aDSS:!SHA1:!AESCCM')
990 context
.minimum_version
= ssl
.TLSVersion
.TLSv1_2
992 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
993 if opts_check_certificate
:
994 if certifi
and 'no-certifi' not in params
.get('compat_opts', []):
995 context
.load_verify_locations(cafile
=certifi
.where())
998 context
.load_default_certs()
999 # Work around the issue in load_default_certs when there are bad certificates. See:
1000 # https://github.com/yt-dlp/yt-dlp/issues/1060,
1001 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
1002 except ssl
.SSLError
:
1003 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
1004 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
1005 for storename
in ('CA', 'ROOT'):
1006 _ssl_load_windows_store_certs(context
, storename
)
1007 context
.set_default_verify_paths()
1009 client_certfile
= params
.get('client_certificate')
1012 context
.load_cert_chain(
1013 client_certfile
, keyfile
=params
.get('client_certificate_key'),
1014 password
=params
.get('client_certificate_password'))
1015 except ssl
.SSLError
:
1016 raise YoutubeDLError('Unable to load client certificate')
1018 # Some servers may reject requests if ALPN extension is not sent. See:
1019 # https://github.com/python/cpython/issues/85140
1020 # https://github.com/yt-dlp/yt-dlp/issues/3878
1021 with contextlib
.suppress(NotImplementedError):
1022 context
.set_alpn_protocols(['http/1.1'])
1024 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
1027 def bug_reports_message(before
=';'):
1028 from ..update
import REPOSITORY
1030 msg
= (f
'please report this issue on https://github.com/{REPOSITORY}/issues?q= , '
1031 'filling out the appropriate issue template. Confirm you are on the latest version using yt-dlp -U')
1033 before
= before
.rstrip()
1034 if not before
or before
.endswith(('.', '!', '?')):
1035 msg
= msg
[0].title() + msg
[1:]
1037 return (before
+ ' ' if before
else '') + msg
1040 class YoutubeDLError(Exception):
1041 """Base exception for YoutubeDL errors."""
1044 def __init__(self
, msg
=None):
1047 elif self
.msg
is None:
1048 self
.msg
= type(self
).__name
__
1049 super().__init
__(self
.msg
)
1052 network_exceptions
= [urllib
.error
.URLError
, http
.client
.HTTPException
, socket
.error
]
1053 if hasattr(ssl
, 'CertificateError'):
1054 network_exceptions
.append(ssl
.CertificateError
)
1055 network_exceptions
= tuple(network_exceptions
)
1058 class ExtractorError(YoutubeDLError
):
1059 """Error during info extraction."""
1061 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
1062 """ tb, if given, is the original traceback (so that it can be printed out).
1063 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
1065 if sys
.exc_info()[0] in network_exceptions
:
1068 self
.orig_msg
= str(msg
)
1070 self
.expected
= expected
1072 self
.video_id
= video_id
1074 self
.exc_info
= sys
.exc_info() # preserve original exception
1075 if isinstance(self
.exc_info
[1], ExtractorError
):
1076 self
.exc_info
= self
.exc_info
[1].exc_info
1077 super().__init
__(self
.__msg
)
1082 format_field(self
.ie
, None, '[%s] '),
1083 format_field(self
.video_id
, None, '%s: '),
1085 format_field(self
.cause
, None, ' (caused by %r)'),
1086 '' if self
.expected
else bug_reports_message()))
1088 def format_traceback(self
):
1089 return join_nonempty(
1090 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
1091 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
1094 def __setattr__(self
, name
, value
):
1095 super().__setattr
__(name
, value
)
1096 if getattr(self
, 'msg', None) and name
not in ('msg', 'args'):
1097 self
.msg
= self
.__msg
or type(self
).__name
__
1098 self
.args
= (self
.msg
, ) # Cannot be property
1101 class UnsupportedError(ExtractorError
):
1102 def __init__(self
, url
):
1104 'Unsupported URL: %s' % url
, expected
=True)
1108 class RegexNotFoundError(ExtractorError
):
1109 """Error when a regex didn't match"""
1113 class GeoRestrictedError(ExtractorError
):
1114 """Geographic restriction Error exception.
1116 This exception may be thrown when a video is not available from your
1117 geographic location due to geographic restrictions imposed by a website.
1120 def __init__(self
, msg
, countries
=None, **kwargs
):
1121 kwargs
['expected'] = True
1122 super().__init
__(msg
, **kwargs
)
1123 self
.countries
= countries
1126 class UserNotLive(ExtractorError
):
1127 """Error when a channel/user is not live"""
1129 def __init__(self
, msg
=None, **kwargs
):
1130 kwargs
['expected'] = True
1131 super().__init
__(msg
or 'The channel is not currently live', **kwargs
)
1134 class DownloadError(YoutubeDLError
):
1135 """Download Error exception.
1137 This exception may be thrown by FileDownloader objects if they are not
1138 configured to continue on errors. They will contain the appropriate
1142 def __init__(self
, msg
, exc_info
=None):
1143 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1144 super().__init
__(msg
)
1145 self
.exc_info
= exc_info
1148 class EntryNotInPlaylist(YoutubeDLError
):
1149 """Entry not in playlist exception.
1151 This exception will be thrown by YoutubeDL when a requested entry
1152 is not found in the playlist info_dict
1154 msg
= 'Entry not found in info'
1157 class SameFileError(YoutubeDLError
):
1158 """Same File exception.
1160 This exception will be thrown by FileDownloader objects if they detect
1161 multiple files would have to be downloaded to the same file on disk.
1163 msg
= 'Fixed output name but more than one file to download'
1165 def __init__(self
, filename
=None):
1166 if filename
is not None:
1167 self
.msg
+= f
': {filename}'
1168 super().__init
__(self
.msg
)
1171 class PostProcessingError(YoutubeDLError
):
1172 """Post Processing exception.
1174 This exception may be raised by PostProcessor's .run() method to
1175 indicate an error in the postprocessing task.
1179 class DownloadCancelled(YoutubeDLError
):
1180 """ Exception raised when the download queue should be interrupted """
1181 msg
= 'The download was cancelled'
1184 class ExistingVideoReached(DownloadCancelled
):
1185 """ --break-on-existing triggered """
1186 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1189 class RejectedVideoReached(DownloadCancelled
):
1190 """ --break-match-filter triggered """
1191 msg
= 'Encountered a video that did not match filter, stopping due to --break-match-filter'
1194 class MaxDownloadsReached(DownloadCancelled
):
1195 """ --max-downloads limit has been reached. """
1196 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1199 class ReExtractInfo(YoutubeDLError
):
1200 """ Video info needs to be re-extracted. """
1202 def __init__(self
, msg
, expected
=False):
1203 super().__init
__(msg
)
1204 self
.expected
= expected
1207 class ThrottledDownload(ReExtractInfo
):
1208 """ Download speed below --throttled-rate. """
1209 msg
= 'The download speed is below throttle limit'
1212 super().__init
__(self
.msg
, expected
=False)
1215 class UnavailableVideoError(YoutubeDLError
):
1216 """Unavailable Format exception.
1218 This exception will be thrown when a video is requested
1219 in a format that is not available for that video.
1221 msg
= 'Unable to download video'
1223 def __init__(self
, err
=None):
1225 self
.msg
+= f
': {err}'
1226 super().__init
__(self
.msg
)
1229 class ContentTooShortError(YoutubeDLError
):
1230 """Content Too Short exception.
1232 This exception may be raised by FileDownloader objects when a file they
1233 download is too small for what the server announced first, indicating
1234 the connection was probably interrupted.
1237 def __init__(self
, downloaded
, expected
):
1238 super().__init
__(f
'Downloaded {downloaded} bytes, expected {expected} bytes')
1240 self
.downloaded
= downloaded
1241 self
.expected
= expected
1244 class XAttrMetadataError(YoutubeDLError
):
1245 def __init__(self
, code
=None, msg
='Unknown error'):
1246 super().__init
__(msg
)
1250 # Parsing code and msg
1251 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1252 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1253 self
.reason
= 'NO_SPACE'
1254 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1255 self
.reason
= 'VALUE_TOO_LONG'
1257 self
.reason
= 'NOT_SUPPORTED'
1260 class XAttrUnavailableError(YoutubeDLError
):
1264 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1265 hc
= http_class(*args
, **kwargs
)
1266 source_address
= ydl_handler
._params
.get('source_address')
1268 if source_address
is not None:
1269 # This is to workaround _create_connection() from socket where it will try all
1270 # address data from getaddrinfo() including IPv6. This filters the result from
1271 # getaddrinfo() based on the source_address value.
1272 # This is based on the cpython socket.create_connection() function.
1273 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1274 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1275 host
, port
= address
1277 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1278 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1279 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1280 if addrs
and not ip_addrs
:
1281 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1283 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1284 % (ip_version
, source_address
[0]))
1285 for res
in ip_addrs
:
1286 af
, socktype
, proto
, canonname
, sa
= res
1289 sock
= socket
.socket(af
, socktype
, proto
)
1290 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1291 sock
.settimeout(timeout
)
1292 sock
.bind(source_address
)
1294 err
= None # Explicitly break reference cycle
1296 except OSError as _
:
1298 if sock
is not None:
1303 raise OSError('getaddrinfo returns an empty list')
1304 if hasattr(hc
, '_create_connection'):
1305 hc
._create
_connection
= _create_connection
1306 hc
.source_address
= (source_address
, 0)
1311 class YoutubeDLHandler(urllib
.request
.HTTPHandler
):
1312 """Handler for HTTP requests and responses.
1314 This class, when installed with an OpenerDirector, automatically adds
1315 the standard headers to every HTTP request and handles gzipped, deflated and
1316 brotli responses from web servers.
1318 Part of this code was copied from:
1320 http://techknack.net/python-urllib2-handlers/
1322 Andrew Rowls, the author of that code, agreed to release it to the
1326 def __init__(self
, params
, *args
, **kwargs
):
1327 urllib
.request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1328 self
._params
= params
1330 def http_open(self
, req
):
1331 conn_class
= http
.client
.HTTPConnection
1333 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1335 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1336 del req
.headers
['Ytdl-socks-proxy']
1338 return self
.do_open(functools
.partial(
1339 _create_http_connection
, self
, conn_class
, False),
1347 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1349 return zlib
.decompress(data
)
1355 return brotli
.decompress(data
)
1357 def http_request(self
, req
):
1358 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1359 # always respected by websites, some tend to give out URLs with non percent-encoded
1360 # non-ASCII characters (see telemb.py, ard.py [#3412])
1361 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1362 # To work around aforementioned issue we will replace request's original URL with
1363 # percent-encoded one
1364 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1365 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1366 url
= req
.get_full_url()
1367 url_escaped
= escape_url(url
)
1369 # Substitute URL if any change after escaping
1370 if url
!= url_escaped
:
1371 req
= update_Request(req
, url
=url_escaped
)
1373 for h
, v
in self
._params
.get('http_headers', std_headers
).items():
1374 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1375 # The dict keys are capitalized because of this bug by urllib
1376 if h
.capitalize() not in req
.headers
:
1377 req
.add_header(h
, v
)
1379 if 'Youtubedl-no-compression' in req
.headers
: # deprecated
1380 req
.headers
.pop('Youtubedl-no-compression', None)
1381 req
.add_header('Accept-encoding', 'identity')
1383 if 'Accept-encoding' not in req
.headers
:
1384 req
.add_header('Accept-encoding', ', '.join(SUPPORTED_ENCODINGS
))
1386 return super().do_request_(req
)
1388 def http_response(self
, req
, resp
):
1391 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1392 content
= resp
.read()
1393 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1395 uncompressed
= io
.BytesIO(gz
.read())
1396 except OSError as original_ioerror
:
1397 # There may be junk add the end of the file
1398 # See http://stackoverflow.com/q/4928560/35070 for details
1399 for i
in range(1, 1024):
1401 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1402 uncompressed
= io
.BytesIO(gz
.read())
1407 raise original_ioerror
1408 resp
= urllib
.request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1409 resp
.msg
= old_resp
.msg
1411 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1412 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1413 resp
= urllib
.request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1414 resp
.msg
= old_resp
.msg
1416 if resp
.headers
.get('Content-encoding', '') == 'br':
1417 resp
= urllib
.request
.addinfourl(
1418 io
.BytesIO(self
.brotli(resp
.read())), old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1419 resp
.msg
= old_resp
.msg
1420 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1421 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1422 if 300 <= resp
.code
< 400:
1423 location
= resp
.headers
.get('Location')
1425 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1426 location
= location
.encode('iso-8859-1').decode()
1427 location_escaped
= escape_url(location
)
1428 if location
!= location_escaped
:
1429 del resp
.headers
['Location']
1430 resp
.headers
['Location'] = location_escaped
1433 https_request
= http_request
1434 https_response
= http_response
1437 def make_socks_conn_class(base_class
, socks_proxy
):
1438 assert issubclass(base_class
, (
1439 http
.client
.HTTPConnection
, http
.client
.HTTPSConnection
))
1441 url_components
= urllib
.parse
.urlparse(socks_proxy
)
1442 if url_components
.scheme
.lower() == 'socks5':
1443 socks_type
= ProxyType
.SOCKS5
1444 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1445 socks_type
= ProxyType
.SOCKS4
1446 elif url_components
.scheme
.lower() == 'socks4a':
1447 socks_type
= ProxyType
.SOCKS4A
1449 def unquote_if_non_empty(s
):
1452 return urllib
.parse
.unquote_plus(s
)
1456 url_components
.hostname
, url_components
.port
or 1080,
1458 unquote_if_non_empty(url_components
.username
),
1459 unquote_if_non_empty(url_components
.password
),
1462 class SocksConnection(base_class
):
1464 self
.sock
= sockssocket()
1465 self
.sock
.setproxy(*proxy_args
)
1466 if isinstance(self
.timeout
, (int, float)):
1467 self
.sock
.settimeout(self
.timeout
)
1468 self
.sock
.connect((self
.host
, self
.port
))
1470 if isinstance(self
, http
.client
.HTTPSConnection
):
1471 if hasattr(self
, '_context'): # Python > 2.6
1472 self
.sock
= self
._context
.wrap_socket(
1473 self
.sock
, server_hostname
=self
.host
)
1475 self
.sock
= ssl
.wrap_socket(self
.sock
)
1477 return SocksConnection
1480 class YoutubeDLHTTPSHandler(urllib
.request
.HTTPSHandler
):
1481 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1482 urllib
.request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1483 self
._https
_conn
_class
= https_conn_class
or http
.client
.HTTPSConnection
1484 self
._params
= params
1486 def https_open(self
, req
):
1488 conn_class
= self
._https
_conn
_class
1490 if hasattr(self
, '_context'): # python > 2.6
1491 kwargs
['context'] = self
._context
1492 if hasattr(self
, '_check_hostname'): # python 3.x
1493 kwargs
['check_hostname'] = self
._check
_hostname
1495 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1497 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1498 del req
.headers
['Ytdl-socks-proxy']
1501 return self
.do_open(
1502 functools
.partial(_create_http_connection
, self
, conn_class
, True), req
, **kwargs
)
1503 except urllib
.error
.URLError
as e
:
1504 if (isinstance(e
.reason
, ssl
.SSLError
)
1505 and getattr(e
.reason
, 'reason', None) == 'SSLV3_ALERT_HANDSHAKE_FAILURE'):
1506 raise YoutubeDLError('SSLV3_ALERT_HANDSHAKE_FAILURE: Try using --legacy-server-connect')
1510 def is_path_like(f
):
1511 return isinstance(f
, (str, bytes, os
.PathLike
))
1514 class YoutubeDLCookieJar(http
.cookiejar
.MozillaCookieJar
):
1516 See [1] for cookie file format.
1518 1. https://curl.haxx.se/docs/http-cookies.html
1520 _HTTPONLY_PREFIX
= '#HttpOnly_'
1522 _HEADER
= '''# Netscape HTTP Cookie File
1523 # This file is generated by yt-dlp. Do not edit.
1526 _CookieFileEntry
= collections
.namedtuple(
1528 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1530 def __init__(self
, filename
=None, *args
, **kwargs
):
1531 super().__init
__(None, *args
, **kwargs
)
1532 if is_path_like(filename
):
1533 filename
= os
.fspath(filename
)
1534 self
.filename
= filename
1537 def _true_or_false(cndn
):
1538 return 'TRUE' if cndn
else 'FALSE'
1540 @contextlib.contextmanager
1541 def open(self
, file, *, write
=False):
1542 if is_path_like(file):
1543 with open(file, 'w' if write
else 'r', encoding
='utf-8') as f
:
1550 def _really_save(self
, f
, ignore_discard
=False, ignore_expires
=False):
1553 if (not ignore_discard
and cookie
.discard
1554 or not ignore_expires
and cookie
.is_expired(now
)):
1556 name
, value
= cookie
.name
, cookie
.value
1558 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1559 # with no name, whereas http.cookiejar regards it as a
1560 # cookie with no value.
1561 name
, value
= '', name
1562 f
.write('%s\n' % '\t'.join((
1564 self
._true
_or
_false
(cookie
.domain
.startswith('.')),
1566 self
._true
_or
_false
(cookie
.secure
),
1567 str_or_none(cookie
.expires
, default
=''),
1571 def save(self
, filename
=None, *args
, **kwargs
):
1573 Save cookies to a file.
1574 Code is taken from CPython 3.6
1575 https://github.com/python/cpython/blob/8d999cbf4adea053be6dbb612b9844635c4dfb8e/Lib/http/cookiejar.py#L2091-L2117 """
1577 if filename
is None:
1578 if self
.filename
is not None:
1579 filename
= self
.filename
1581 raise ValueError(http
.cookiejar
.MISSING_FILENAME_TEXT
)
1583 # Store session cookies with `expires` set to 0 instead of an empty string
1585 if cookie
.expires
is None:
1588 with self
.open(filename
, write
=True) as f
:
1589 f
.write(self
._HEADER
)
1590 self
._really
_save
(f
, *args
, **kwargs
)
1592 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1593 """Load cookies from a file."""
1594 if filename
is None:
1595 if self
.filename
is not None:
1596 filename
= self
.filename
1598 raise ValueError(http
.cookiejar
.MISSING_FILENAME_TEXT
)
1600 def prepare_line(line
):
1601 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1602 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1603 # comments and empty lines are fine
1604 if line
.startswith('#') or not line
.strip():
1606 cookie_list
= line
.split('\t')
1607 if len(cookie_list
) != self
._ENTRY
_LEN
:
1608 raise http
.cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1609 cookie
= self
._CookieFileEntry
(*cookie_list
)
1610 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1611 raise http
.cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1615 with self
.open(filename
) as f
:
1618 cf
.write(prepare_line(line
))
1619 except http
.cookiejar
.LoadError
as e
:
1620 if f
'{line.strip()} '[0] in '[{"':
1621 raise http
.cookiejar
.LoadError(
1622 'Cookies file must be Netscape formatted, not JSON. See '
1623 'https://github.com/yt-dlp/yt-dlp/wiki/FAQ#how-do-i-pass-cookies-to-yt-dlp')
1624 write_string(f
'WARNING: skipping cookie file entry due to {e}: {line!r}\n')
1627 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1628 # Session cookies are denoted by either `expires` field set to
1629 # an empty string or 0. MozillaCookieJar only recognizes the former
1630 # (see [1]). So we need force the latter to be recognized as session
1631 # cookies on our own.
1632 # Session cookies may be important for cookies-based authentication,
1633 # e.g. usually, when user does not check 'Remember me' check box while
1634 # logging in on a site, some important cookies are stored as session
1635 # cookies so that not recognizing them will result in failed login.
1636 # 1. https://bugs.python.org/issue17164
1638 # Treat `expires=0` cookies as session cookies
1639 if cookie
.expires
== 0:
1640 cookie
.expires
= None
1641 cookie
.discard
= True
1644 class YoutubeDLCookieProcessor(urllib
.request
.HTTPCookieProcessor
):
1645 def __init__(self
, cookiejar
=None):
1646 urllib
.request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1648 def http_response(self
, request
, response
):
1649 return urllib
.request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1651 https_request
= urllib
.request
.HTTPCookieProcessor
.http_request
1652 https_response
= http_response
1655 class YoutubeDLRedirectHandler(urllib
.request
.HTTPRedirectHandler
):
1656 """YoutubeDL redirect handler
1658 The code is based on HTTPRedirectHandler implementation from CPython [1].
1660 This redirect handler solves two issues:
1661 - ensures redirect URL is always unicode under python 2
1662 - introduces support for experimental HTTP response status code
1663 308 Permanent Redirect [2] used by some sites [3]
1665 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1666 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1667 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1670 http_error_301
= http_error_303
= http_error_307
= http_error_308
= urllib
.request
.HTTPRedirectHandler
.http_error_302
1672 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1673 """Return a Request or None in response to a redirect.
1675 This is called by the http_error_30x methods when a
1676 redirection response is received. If a redirection should
1677 take place, return a new Request to allow http_error_30x to
1678 perform the redirect. Otherwise, raise HTTPError if no-one
1679 else should try to handle this url. Return None if you can't
1680 but another Handler might.
1682 m
= req
.get_method()
1683 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1684 or code
in (301, 302, 303) and m
== "POST")):
1685 raise urllib
.error
.HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1686 # Strictly (according to RFC 2616), 301 or 302 in response to
1687 # a POST MUST NOT cause a redirection without confirmation
1688 # from the user (of urllib.request, in this case). In practice,
1689 # essentially all clients do redirect in this case, so we do
1692 # Be conciliant with URIs containing a space. This is mainly
1693 # redundant with the more complete encoding done in http_error_302(),
1694 # but it is kept for compatibility with other callers.
1695 newurl
= newurl
.replace(' ', '%20')
1697 CONTENT_HEADERS
= ("content-length", "content-type")
1698 # NB: don't use dict comprehension for python 2.6 compatibility
1699 newheaders
= {k: v for k, v in req.headers.items() if k.lower() not in CONTENT_HEADERS}
1701 # A 303 must either use GET or HEAD for subsequent request
1702 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.4
1703 if code
== 303 and m
!= 'HEAD':
1705 # 301 and 302 redirects are commonly turned into a GET from a POST
1706 # for subsequent requests by browsers, so we'll do the same.
1707 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.2
1708 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.3
1709 if code
in (301, 302) and m
== 'POST':
1712 return urllib
.request
.Request(
1713 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1714 unverifiable
=True, method
=m
)
1717 def extract_timezone(date_str
):
1720 ^.{8,}? # >=8 char non-TZ prefix, if present
1721 (?P<tz>Z| # just the UTC Z, or
1722 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1723 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1724 [ ]? # optional space
1725 (?P<sign>\+|-) # +/-
1726 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1730 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1731 timezone
= TIMEZONE_NAMES
.get(m
and m
.group('tz').strip())
1732 if timezone
is not None:
1733 date_str
= date_str
[:-len(m
.group('tz'))]
1734 timezone
= datetime
.timedelta(hours
=timezone
or 0)
1736 date_str
= date_str
[:-len(m
.group('tz'))]
1737 if not m
.group('sign'):
1738 timezone
= datetime
.timedelta()
1740 sign
= 1 if m
.group('sign') == '+' else -1
1741 timezone
= datetime
.timedelta(
1742 hours
=sign
* int(m
.group('hours')),
1743 minutes
=sign
* int(m
.group('minutes')))
1744 return timezone
, date_str
1747 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1748 """ Return a UNIX timestamp from the given date """
1750 if date_str
is None:
1753 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1755 if timezone
is None:
1756 timezone
, date_str
= extract_timezone(date_str
)
1758 with contextlib
.suppress(ValueError):
1759 date_format
= f
'%Y-%m-%d{delimiter}%H:%M:%S'
1760 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1761 return calendar
.timegm(dt
.timetuple())
1764 def date_formats(day_first
=True):
1765 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1768 def unified_strdate(date_str
, day_first
=True):
1769 """Return a string with the date in the format YYYYMMDD"""
1771 if date_str
is None:
1775 date_str
= date_str
.replace(',', ' ')
1776 # Remove AM/PM + timezone
1777 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1778 _
, date_str
= extract_timezone(date_str
)
1780 for expression
in date_formats(day_first
):
1781 with contextlib
.suppress(ValueError):
1782 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1783 if upload_date
is None:
1784 timetuple
= email
.utils
.parsedate_tz(date_str
)
1786 with contextlib
.suppress(ValueError):
1787 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1788 if upload_date
is not None:
1789 return str(upload_date
)
1792 def unified_timestamp(date_str
, day_first
=True):
1793 if date_str
is None:
1796 date_str
= re
.sub(r
'\s+', ' ', re
.sub(
1797 r
'(?i)[,|]|(mon|tues?|wed(nes)?|thu(rs)?|fri|sat(ur)?)(day)?', '', date_str
))
1799 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1800 timezone
, date_str
= extract_timezone(date_str
)
1802 # Remove AM/PM + timezone
1803 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1805 # Remove unrecognized timezones from ISO 8601 alike timestamps
1806 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1808 date_str
= date_str
[:-len(m
.group('tz'))]
1810 # Python only supports microseconds, so remove nanoseconds
1811 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
1813 date_str
= m
.group(1)
1815 for expression
in date_formats(day_first
):
1816 with contextlib
.suppress(ValueError):
1817 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1818 return calendar
.timegm(dt
.timetuple())
1820 timetuple
= email
.utils
.parsedate_tz(date_str
)
1822 return calendar
.timegm(timetuple
) + pm_delta
* 3600 - timezone
.total_seconds()
1825 def determine_ext(url
, default_ext
='unknown_video'):
1826 if url
is None or '.' not in url
:
1828 guess
= url
.partition('?')[0].rpartition('.')[2]
1829 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1831 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1832 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1833 return guess
.rstrip('/')
1838 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1839 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1842 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1844 Return a datetime object from a string.
1846 (now|today|yesterday|DATE)([+-]\d+(microsecond|second|minute|hour|day|week|month|year)s?)?
1848 @param format strftime format of DATE
1849 @param precision Round the datetime object: auto|microsecond|second|minute|hour|day
1850 auto: round to the unit provided in date_str (if applicable).
1852 auto_precision
= False
1853 if precision
== 'auto':
1854 auto_precision
= True
1855 precision
= 'microsecond'
1856 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1857 if date_str
in ('now', 'today'):
1859 if date_str
== 'yesterday':
1860 return today
- datetime
.timedelta(days
=1)
1862 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)s?',
1864 if match
is not None:
1865 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1866 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1867 unit
= match
.group('unit')
1868 if unit
== 'month' or unit
== 'year':
1869 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1875 delta
= datetime
.timedelta(**{unit + 's': time}
)
1876 new_date
= start_time
+ delta
1878 return datetime_round(new_date
, unit
)
1881 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1884 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1886 Return a date object from a string using datetime_from_str
1888 @param strict Restrict allowed patterns to "YYYYMMDD" and
1889 (now|today|yesterday)(-\d+(day|week|month|year)s?)?
1891 if strict
and not re
.fullmatch(r
'\d{8}|(now|today|yesterday)(-\d+(day|week|month|year)s?)?', date_str
):
1892 raise ValueError(f
'Invalid date format "{date_str}"')
1893 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1896 def datetime_add_months(dt
, months
):
1897 """Increment/Decrement a datetime object by months."""
1898 month
= dt
.month
+ months
- 1
1899 year
= dt
.year
+ month
// 12
1900 month
= month
% 12 + 1
1901 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1902 return dt
.replace(year
, month
, day
)
1905 def datetime_round(dt
, precision
='day'):
1907 Round a datetime object's time to a specific precision
1909 if precision
== 'microsecond':
1918 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1919 timestamp
= calendar
.timegm(dt
.timetuple())
1920 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1923 def hyphenate_date(date_str
):
1925 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1926 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1927 if match
is not None:
1928 return '-'.join(match
.groups())
1934 """Represents a time interval between two dates"""
1936 def __init__(self
, start
=None, end
=None):
1937 """start and end must be strings in the format accepted by date"""
1938 if start
is not None:
1939 self
.start
= date_from_str(start
, strict
=True)
1941 self
.start
= datetime
.datetime
.min.date()
1943 self
.end
= date_from_str(end
, strict
=True)
1945 self
.end
= datetime
.datetime
.max.date()
1946 if self
.start
> self
.end
:
1947 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1951 """Returns a range that only contains the given day"""
1952 return cls(day
, day
)
1954 def __contains__(self
, date
):
1955 """Check if the date is in the range"""
1956 if not isinstance(date
, datetime
.date
):
1957 date
= date_from_str(date
)
1958 return self
.start
<= date
<= self
.end
1961 return f
'{self.start.isoformat()} - {self.end.isoformat()}'
1963 def __eq__(self
, other
):
1964 return (isinstance(other
, DateRange
)
1965 and self
.start
== other
.start
and self
.end
== other
.end
)
1969 def system_identifier():
1970 python_implementation
= platform
.python_implementation()
1971 if python_implementation
== 'PyPy' and hasattr(sys
, 'pypy_version_info'):
1972 python_implementation
+= ' version %d.%d.%d' % sys
.pypy_version_info
[:3]
1974 with contextlib
.suppress(OSError): # We may not have access to the executable
1975 libc_ver
= platform
.libc_ver()
1977 return 'Python %s (%s %s %s) - %s (%s%s)' % (
1978 platform
.python_version(),
1979 python_implementation
,
1981 platform
.architecture()[0],
1982 platform
.platform(),
1983 ssl
.OPENSSL_VERSION
,
1984 format_field(join_nonempty(*libc_ver
, delim
=' '), None, ', %s'),
1989 def get_windows_version():
1990 ''' Get Windows version. returns () if it's not running on Windows '''
1991 if compat_os_name
== 'nt':
1992 return version_tuple(platform
.win32_ver()[1])
1997 def write_string(s
, out
=None, encoding
=None):
1998 assert isinstance(s
, str)
1999 out
= out
or sys
.stderr
2000 # `sys.stderr` might be `None` (Ref: https://github.com/pyinstaller/pyinstaller/pull/7217)
2004 if compat_os_name
== 'nt' and supports_terminal_sequences(out
):
2005 s
= re
.sub(r
'([\r\n]+)', r
' \1', s
)
2007 enc
, buffer = None, out
2008 if 'b' in getattr(out
, 'mode', ''):
2009 enc
= encoding
or preferredencoding()
2010 elif hasattr(out
, 'buffer'):
2012 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
2014 buffer.write(s
.encode(enc
, 'ignore') if enc
else s
)
2018 def deprecation_warning(msg
, *, printer
=None, stacklevel
=0, **kwargs
):
2019 from .. import _IN_CLI
2021 if msg
in deprecation_warning
._cache
:
2023 deprecation_warning
._cache
.add(msg
)
2025 return printer(f
'{msg}{bug_reports_message()}', **kwargs
)
2026 return write_string(f
'ERROR: {msg}{bug_reports_message()}\n', **kwargs
)
2029 warnings
.warn(DeprecationWarning(msg
), stacklevel
=stacklevel
+ 3)
2032 deprecation_warning
._cache
= set()
2035 def bytes_to_intlist(bs
):
2038 if isinstance(bs
[0], int): # Python 3
2041 return [ord(c
) for c
in bs
]
2044 def intlist_to_bytes(xs
):
2047 return struct
.pack('%dB' % len(xs
), *xs
)
2050 class LockingUnsupportedError(OSError):
2051 msg
= 'File locking is not supported'
2054 super().__init
__(self
.msg
)
2057 # Cross-platform file locking
2058 if sys
.platform
== 'win32':
2060 import ctypes
.wintypes
2063 class OVERLAPPED(ctypes
.Structure
):
2065 ('Internal', ctypes
.wintypes
.LPVOID
),
2066 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
2067 ('Offset', ctypes
.wintypes
.DWORD
),
2068 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
2069 ('hEvent', ctypes
.wintypes
.HANDLE
),
2072 kernel32
= ctypes
.WinDLL('kernel32')
2073 LockFileEx
= kernel32
.LockFileEx
2074 LockFileEx
.argtypes
= [
2075 ctypes
.wintypes
.HANDLE
, # hFile
2076 ctypes
.wintypes
.DWORD
, # dwFlags
2077 ctypes
.wintypes
.DWORD
, # dwReserved
2078 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2079 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2080 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2082 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
2083 UnlockFileEx
= kernel32
.UnlockFileEx
2084 UnlockFileEx
.argtypes
= [
2085 ctypes
.wintypes
.HANDLE
, # hFile
2086 ctypes
.wintypes
.DWORD
, # dwReserved
2087 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2088 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2089 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2091 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
2092 whole_low
= 0xffffffff
2093 whole_high
= 0x7fffffff
2095 def _lock_file(f
, exclusive
, block
):
2096 overlapped
= OVERLAPPED()
2097 overlapped
.Offset
= 0
2098 overlapped
.OffsetHigh
= 0
2099 overlapped
.hEvent
= 0
2100 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
2102 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
2103 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
2104 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2105 # NB: No argument form of "ctypes.FormatError" does not work on PyPy
2106 raise BlockingIOError(f
'Locking file failed: {ctypes.FormatError(ctypes.GetLastError())!r}')
2108 def _unlock_file(f
):
2109 assert f
._lock
_file
_overlapped
_p
2110 handle
= msvcrt
.get_osfhandle(f
.fileno())
2111 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2112 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
2118 def _lock_file(f
, exclusive
, block
):
2119 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
2121 flags |
= fcntl
.LOCK_NB
2123 fcntl
.flock(f
, flags
)
2124 except BlockingIOError
:
2126 except OSError: # AOSP does not have flock()
2127 fcntl
.lockf(f
, flags
)
2129 def _unlock_file(f
):
2130 with contextlib
.suppress(OSError):
2131 return fcntl
.flock(f
, fcntl
.LOCK_UN
)
2132 with contextlib
.suppress(OSError):
2133 return fcntl
.lockf(f
, fcntl
.LOCK_UN
) # AOSP does not have flock()
2134 return fcntl
.flock(f
, fcntl
.LOCK_UN | fcntl
.LOCK_NB
) # virtiofs needs LOCK_NB on unlocking
2138 def _lock_file(f
, exclusive
, block
):
2139 raise LockingUnsupportedError()
2141 def _unlock_file(f
):
2142 raise LockingUnsupportedError()
2148 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
2149 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
2150 raise NotImplementedError(mode
)
2151 self
.mode
, self
.block
= mode
, block
2153 writable
= any(f
in mode
for f
in 'wax+')
2154 readable
= any(f
in mode
for f
in 'r+')
2155 flags
= functools
.reduce(operator
.ior
, (
2156 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
2157 getattr(os
, 'O_BINARY', 0), # Windows only
2158 getattr(os
, 'O_NOINHERIT', 0), # Windows only
2159 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
2160 os
.O_APPEND
if 'a' in mode
else 0,
2161 os
.O_EXCL
if 'x' in mode
else 0,
2162 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
2165 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
2167 def __enter__(self
):
2168 exclusive
= 'r' not in self
.mode
2170 _lock_file(self
.f
, exclusive
, self
.block
)
2175 if 'w' in self
.mode
:
2178 except OSError as e
:
2180 errno
.ESPIPE
, # Illegal seek - expected for FIFO
2181 errno
.EINVAL
, # Invalid argument - expected for /dev/null
2190 _unlock_file(self
.f
)
2194 def __exit__(self
, *_
):
2203 def __getattr__(self
, attr
):
2204 return getattr(self
.f
, attr
)
2211 def get_filesystem_encoding():
2212 encoding
= sys
.getfilesystemencoding()
2213 return encoding
if encoding
is not None else 'utf-8'
2216 def shell_quote(args
):
2218 encoding
= get_filesystem_encoding()
2220 if isinstance(a
, bytes):
2221 # We may get a filename encoded with 'encodeFilename'
2222 a
= a
.decode(encoding
)
2223 quoted_args
.append(compat_shlex_quote(a
))
2224 return ' '.join(quoted_args
)
2227 def smuggle_url(url
, data
):
2228 """ Pass additional data in a URL for internal use. """
2230 url
, idata
= unsmuggle_url(url
, {})
2232 sdata
= urllib
.parse
.urlencode(
2233 {'__youtubedl_smuggle': json.dumps(data)}
)
2234 return url
+ '#' + sdata
2237 def unsmuggle_url(smug_url
, default
=None):
2238 if '#__youtubedl_smuggle' not in smug_url
:
2239 return smug_url
, default
2240 url
, _
, sdata
= smug_url
.rpartition('#')
2241 jsond
= urllib
.parse
.parse_qs(sdata
)['__youtubedl_smuggle'][0]
2242 data
= json
.loads(jsond
)
2246 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2247 """ Formats numbers with decimal sufixes like K, M, etc """
2248 num
, factor
= float_or_none(num
), float(factor
)
2249 if num
is None or num
< 0:
2251 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
2252 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
2253 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
2255 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2256 converted
= num
/ (factor
** exponent
)
2257 return fmt
% (converted
, suffix
)
2260 def format_bytes(bytes):
2261 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2264 def lookup_unit_table(unit_table
, s
, strict
=False):
2265 num_re
= NUMBER_RE
if strict
else NUMBER_RE
.replace(R
'\.', '[,.]')
2266 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2267 m
= (re
.fullmatch
if strict
else re
.match
)(
2268 rf
'(?P<num>{num_re})\s*(?P<unit>{units_re})\b', s
)
2272 num
= float(m
.group('num').replace(',', '.'))
2273 mult
= unit_table
[m
.group('unit')]
2274 return round(num
* mult
)
2278 """Parse a string indicating a byte quantity into an integer"""
2279 return lookup_unit_table(
2280 {u: 1024**i for i, u in enumerate(['', *'KMGTPEZY'])}
,
2281 s
.upper(), strict
=True)
2284 def parse_filesize(s
):
2288 # The lower-case forms are of course incorrect and unofficial,
2289 # but we support those too
2306 'megabytes': 1000 ** 2,
2307 'mebibytes': 1024 ** 2,
2313 'gigabytes': 1000 ** 3,
2314 'gibibytes': 1024 ** 3,
2320 'terabytes': 1000 ** 4,
2321 'tebibytes': 1024 ** 4,
2327 'petabytes': 1000 ** 5,
2328 'pebibytes': 1024 ** 5,
2334 'exabytes': 1000 ** 6,
2335 'exbibytes': 1024 ** 6,
2341 'zettabytes': 1000 ** 7,
2342 'zebibytes': 1024 ** 7,
2348 'yottabytes': 1000 ** 8,
2349 'yobibytes': 1024 ** 8,
2352 return lookup_unit_table(_UNIT_TABLE
, s
)
2359 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2361 if re
.match(r
'^[\d,.]+$', s
):
2362 return str_to_int(s
)
2375 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2379 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2381 return str_to_int(mobj
.group(1))
2384 def parse_resolution(s
, *, lenient
=False):
2389 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
2391 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2394 'width': int(mobj
.group('w')),
2395 'height': int(mobj
.group('h')),
2398 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2400 return {'height': int(mobj.group(1))}
2402 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2404 return {'height': int(mobj.group(1)) * 540}
2409 def parse_bitrate(s
):
2410 if not isinstance(s
, str):
2412 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2414 return int(mobj
.group(1))
2417 def month_by_name(name
, lang
='en'):
2418 """ Return the number of a month by (locale-independently) English name """
2420 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2423 return month_names
.index(name
) + 1
2428 def month_by_abbreviation(abbrev
):
2429 """ Return the number of a month by (locale-independently) English
2433 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2438 def fix_xml_ampersands(xml_str
):
2439 """Replace all the '&' by '&' in XML"""
2441 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2446 def setproctitle(title
):
2447 assert isinstance(title
, str)
2449 # Workaround for https://github.com/yt-dlp/yt-dlp/issues/4541
2456 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2460 # LoadLibrary in Windows Python 2.7.13 only expects
2461 # a bytestring, but since unicode_literals turns
2462 # every string into a unicode string, it fails.
2464 title_bytes
= title
.encode()
2465 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2466 buf
.value
= title_bytes
2468 libc
.prctl(15, buf
, 0, 0, 0)
2469 except AttributeError:
2470 return # Strange libc, just skip this
2473 def remove_start(s
, start
):
2474 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2477 def remove_end(s
, end
):
2478 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2481 def remove_quotes(s
):
2482 if s
is None or len(s
) < 2:
2484 for quote
in ('"', "'", ):
2485 if s
[0] == quote
and s
[-1] == quote
:
2490 def get_domain(url
):
2492 This implementation is inconsistent, but is kept for compatibility.
2493 Use this only for "webpage_url_domain"
2495 return remove_start(urllib
.parse
.urlparse(url
).netloc
, 'www.') or None
2498 def url_basename(url
):
2499 path
= urllib
.parse
.urlparse(url
).path
2500 return path
.strip('/').split('/')[-1]
2504 return re
.match(r
'https?://[^?#]+/', url
).group()
2507 def urljoin(base
, path
):
2508 if isinstance(path
, bytes):
2509 path
= path
.decode()
2510 if not isinstance(path
, str) or not path
:
2512 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2514 if isinstance(base
, bytes):
2515 base
= base
.decode()
2516 if not isinstance(base
, str) or not re
.match(
2517 r
'^(?:https?:)?//', base
):
2519 return urllib
.parse
.urljoin(base
, path
)
2522 class HEADRequest(urllib
.request
.Request
):
2523 def get_method(self
):
2527 class PUTRequest(urllib
.request
.Request
):
2528 def get_method(self
):
2532 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2533 if get_attr
and v
is not None:
2534 v
= getattr(v
, get_attr
, None)
2536 return int(v
) * invscale
// scale
2537 except (ValueError, TypeError, OverflowError):
2541 def str_or_none(v
, default
=None):
2542 return default
if v
is None else str(v
)
2545 def str_to_int(int_str
):
2546 """ A more relaxed version of int_or_none """
2547 if isinstance(int_str
, int):
2549 elif isinstance(int_str
, str):
2550 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2551 return int_or_none(int_str
)
2554 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2558 return float(v
) * invscale
/ scale
2559 except (ValueError, TypeError):
2563 def bool_or_none(v
, default
=None):
2564 return v
if isinstance(v
, bool) else default
2567 def strip_or_none(v
, default
=None):
2568 return v
.strip() if isinstance(v
, str) else default
2571 def url_or_none(url
):
2572 if not url
or not isinstance(url
, str):
2575 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2578 def request_to_url(req
):
2579 if isinstance(req
, urllib
.request
.Request
):
2580 return req
.get_full_url()
2585 def strftime_or_none(timestamp
, date_format
, default
=None):
2586 datetime_object
= None
2588 if isinstance(timestamp
, (int, float)): # unix timestamp
2589 # Using naive datetime here can break timestamp() in Windows
2590 # Ref: https://github.com/yt-dlp/yt-dlp/issues/5185, https://github.com/python/cpython/issues/94414
2591 datetime_object
= datetime
.datetime
.fromtimestamp(timestamp
, datetime
.timezone
.utc
)
2592 elif isinstance(timestamp
, str): # assume YYYYMMDD
2593 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2594 date_format
= re
.sub( # Support %s on windows
2595 r
'(?<!%)(%%)*%s', rf
'\g<1>{int(datetime_object.timestamp())}', date_format
)
2596 return datetime_object
.strftime(date_format
)
2597 except (ValueError, TypeError, AttributeError):
2601 def parse_duration(s
):
2602 if not isinstance(s
, str):
2608 days
, hours
, mins
, secs
, ms
= [None] * 5
2609 m
= re
.match(r
'''(?x)
2611 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2612 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2613 (?P<ms>[.:][0-9]+)?Z?$
2616 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2621 [0-9]+\s*y(?:ears?)?,?\s*
2624 [0-9]+\s*m(?:onths?)?,?\s*
2627 [0-9]+\s*w(?:eeks?)?,?\s*
2630 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2634 (?P<hours>[0-9]+)\s*h(?:ours?)?,?\s*
2637 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2640 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2643 days
, hours
, mins
, secs
, ms
= m
.groups()
2645 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2647 hours
, mins
= m
.groups()
2652 ms
= ms
.replace(':', '.')
2653 return sum(float(part
or 0) * mult
for part
, mult
in (
2654 (days
, 86400), (hours
, 3600), (mins
, 60), (secs
, 1), (ms
, 1)))
2657 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2658 name
, real_ext
= os
.path
.splitext(filename
)
2660 f
'{name}.{ext}{real_ext}'
2661 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2662 else f
'{filename}.{ext}')
2665 def replace_extension(filename
, ext
, expected_real_ext
=None):
2666 name
, real_ext
= os
.path
.splitext(filename
)
2667 return '{}.{}'.format(
2668 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2672 def check_executable(exe
, args
=[]):
2673 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2674 args can be a list of arguments for a short output (like -version) """
2676 Popen
.run([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
)
2682 def _get_exe_version_output(exe
, args
):
2684 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2685 # SIGTTOU if yt-dlp is run in the background.
2686 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2687 stdout
, _
, ret
= Popen
.run([encodeArgument(exe
)] + args
, text
=True,
2688 stdin
=subprocess
.PIPE
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
)
2696 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2697 assert isinstance(output
, str)
2698 if version_re
is None:
2699 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2700 m
= re
.search(version_re
, output
)
2707 def get_exe_version(exe
, args
=['--version'],
2708 version_re
=None, unrecognized
=('present', 'broken')):
2709 """ Returns the version of the specified executable,
2710 or False if the executable is not present """
2711 unrecognized
= variadic(unrecognized
)
2712 assert len(unrecognized
) in (1, 2)
2713 out
= _get_exe_version_output(exe
, args
)
2715 return unrecognized
[-1]
2716 return out
and detect_exe_version(out
, version_re
, unrecognized
[0])
2719 def frange(start
=0, stop
=None, step
=1):
2722 start
, stop
= 0, start
2723 sign
= [-1, 1][step
> 0] if step
else 0
2724 while sign
* start
< sign
* stop
:
2729 class LazyList(collections
.abc
.Sequence
):
2730 """Lazy immutable list from an iterable
2731 Note that slices of a LazyList are lists and not LazyList"""
2733 class IndexError(IndexError):
2736 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2737 self
._iterable
= iter(iterable
)
2738 self
._cache
= [] if _cache
is None else _cache
2739 self
._reversed
= reverse
2743 # We need to consume the entire iterable to iterate in reverse
2744 yield from self
.exhaust()
2746 yield from self
._cache
2747 for item
in self
._iterable
:
2748 self
._cache
.append(item
)
2752 self
._cache
.extend(self
._iterable
)
2753 self
._iterable
= [] # Discard the emptied iterable to make it pickle-able
2757 """Evaluate the entire iterable"""
2758 return self
._exhaust
()[::-1 if self
._reversed
else 1]
2761 def _reverse_index(x
):
2762 return None if x
is None else ~x
2764 def __getitem__(self
, idx
):
2765 if isinstance(idx
, slice):
2767 idx
= slice(self
._reverse
_index
(idx
.start
), self
._reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2768 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2769 elif isinstance(idx
, int):
2771 idx
= self
._reverse
_index
(idx
)
2772 start
, stop
, step
= idx
, idx
, 0
2774 raise TypeError('indices must be integers or slices')
2775 if ((start
or 0) < 0 or (stop
or 0) < 0
2776 or (start
is None and step
< 0)
2777 or (stop
is None and step
> 0)):
2778 # We need to consume the entire iterable to be able to slice from the end
2779 # Obviously, never use this with infinite iterables
2782 return self
._cache
[idx
]
2783 except IndexError as e
:
2784 raise self
.IndexError(e
) from e
2785 n
= max(start
or 0, stop
or 0) - len(self
._cache
) + 1
2787 self
._cache
.extend(itertools
.islice(self
._iterable
, n
))
2789 return self
._cache
[idx
]
2790 except IndexError as e
:
2791 raise self
.IndexError(e
) from e
2795 self
[-1] if self
._reversed
else self
[0]
2796 except self
.IndexError:
2802 return len(self
._cache
)
2804 def __reversed__(self
):
2805 return type(self
)(self
._iterable
, reverse
=not self
._reversed
, _cache
=self
._cache
)
2808 return type(self
)(self
._iterable
, reverse
=self
._reversed
, _cache
=self
._cache
)
2811 # repr and str should mimic a list. So we exhaust the iterable
2812 return repr(self
.exhaust())
2815 return repr(self
.exhaust())
2820 class IndexError(IndexError):
2824 # This is only useful for tests
2825 return len(self
.getslice())
2827 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2828 self
._pagefunc
= pagefunc
2829 self
._pagesize
= pagesize
2830 self
._pagecount
= float('inf')
2831 self
._use
_cache
= use_cache
2834 def getpage(self
, pagenum
):
2835 page_results
= self
._cache
.get(pagenum
)
2836 if page_results
is None:
2837 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2839 self
._cache
[pagenum
] = page_results
2842 def getslice(self
, start
=0, end
=None):
2843 return list(self
._getslice
(start
, end
))
2845 def _getslice(self
, start
, end
):
2846 raise NotImplementedError('This method must be implemented by subclasses')
2848 def __getitem__(self
, idx
):
2849 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2850 if not isinstance(idx
, int) or idx
< 0:
2851 raise TypeError('indices must be non-negative integers')
2852 entries
= self
.getslice(idx
, idx
+ 1)
2854 raise self
.IndexError()
2858 class OnDemandPagedList(PagedList
):
2859 """Download pages until a page with less than maximum results"""
2861 def _getslice(self
, start
, end
):
2862 for pagenum
in itertools
.count(start
// self
._pagesize
):
2863 firstid
= pagenum
* self
._pagesize
2864 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2865 if start
>= nextfirstid
:
2869 start
% self
._pagesize
2870 if firstid
<= start
< nextfirstid
2873 ((end
- 1) % self
._pagesize
) + 1
2874 if (end
is not None and firstid
<= end
<= nextfirstid
)
2878 page_results
= self
.getpage(pagenum
)
2880 self
._pagecount
= pagenum
- 1
2882 if startv
!= 0 or endv
is not None:
2883 page_results
= page_results
[startv
:endv
]
2884 yield from page_results
2886 # A little optimization - if current page is not "full", ie. does
2887 # not contain page_size videos then we can assume that this page
2888 # is the last one - there are no more ids on further pages -
2889 # i.e. no need to query again.
2890 if len(page_results
) + startv
< self
._pagesize
:
2893 # If we got the whole page, but the next page is not interesting,
2894 # break out early as well
2895 if end
== nextfirstid
:
2899 class InAdvancePagedList(PagedList
):
2900 """PagedList with total number of pages known in advance"""
2902 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2903 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2904 self
._pagecount
= pagecount
2906 def _getslice(self
, start
, end
):
2907 start_page
= start
// self
._pagesize
2908 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2909 skip_elems
= start
- start_page
* self
._pagesize
2910 only_more
= None if end
is None else end
- start
2911 for pagenum
in range(start_page
, end_page
):
2912 page_results
= self
.getpage(pagenum
)
2914 page_results
= page_results
[skip_elems
:]
2916 if only_more
is not None:
2917 if len(page_results
) < only_more
:
2918 only_more
-= len(page_results
)
2920 yield from page_results
[:only_more
]
2922 yield from page_results
2925 class PlaylistEntries
:
2926 MissingEntry
= object()
2927 is_exhausted
= False
2929 def __init__(self
, ydl
, info_dict
):
2932 # _entries must be assigned now since infodict can change during iteration
2933 entries
= info_dict
.get('entries')
2935 raise EntryNotInPlaylist('There are no entries')
2936 elif isinstance(entries
, list):
2937 self
.is_exhausted
= True
2939 requested_entries
= info_dict
.get('requested_entries')
2940 self
.is_incomplete
= requested_entries
is not None
2941 if self
.is_incomplete
:
2942 assert self
.is_exhausted
2943 self
._entries
= [self
.MissingEntry
] * max(requested_entries
or [0])
2944 for i
, entry
in zip(requested_entries
, entries
):
2945 self
._entries
[i
- 1] = entry
2946 elif isinstance(entries
, (list, PagedList
, LazyList
)):
2947 self
._entries
= entries
2949 self
._entries
= LazyList(entries
)
2951 PLAYLIST_ITEMS_RE
= re
.compile(r
'''(?x)
2952 (?P<start>[+-]?\d+)?
2954 (?P<end>[+-]?\d+|inf(?:inite)?)?
2955 (?::(?P<step>[+-]?\d+))?
2959 def parse_playlist_items(cls
, string
):
2960 for segment
in string
.split(','):
2962 raise ValueError('There is two or more consecutive commas')
2963 mobj
= cls
.PLAYLIST_ITEMS_RE
.fullmatch(segment
)
2965 raise ValueError(f
'{segment!r} is not a valid specification')
2966 start
, end
, step
, has_range
= mobj
.group('start', 'end', 'step', 'range')
2967 if int_or_none(step
) == 0:
2968 raise ValueError(f
'Step in {segment!r} cannot be zero')
2969 yield slice(int_or_none(start
), float_or_none(end
), int_or_none(step
)) if has_range
else int(start
)
2971 def get_requested_items(self
):
2972 playlist_items
= self
.ydl
.params
.get('playlist_items')
2973 playlist_start
= self
.ydl
.params
.get('playliststart', 1)
2974 playlist_end
= self
.ydl
.params
.get('playlistend')
2975 # For backwards compatibility, interpret -1 as whole list
2976 if playlist_end
in (-1, None):
2978 if not playlist_items
:
2979 playlist_items
= f
'{playlist_start}:{playlist_end}'
2980 elif playlist_start
!= 1 or playlist_end
:
2981 self
.ydl
.report_warning('Ignoring playliststart and playlistend because playlistitems was given', only_once
=True)
2983 for index
in self
.parse_playlist_items(playlist_items
):
2984 for i
, entry
in self
[index
]:
2989 # The item may have just been added to archive. Don't break due to it
2990 if not self
.ydl
.params
.get('lazy_playlist'):
2991 # TODO: Add auto-generated fields
2992 self
.ydl
._match
_entry
(entry
, incomplete
=True, silent
=True)
2993 except (ExistingVideoReached
, RejectedVideoReached
):
2996 def get_full_count(self
):
2997 if self
.is_exhausted
and not self
.is_incomplete
:
2999 elif isinstance(self
._entries
, InAdvancePagedList
):
3000 if self
._entries
._pagesize
== 1:
3001 return self
._entries
._pagecount
3003 @functools.cached_property
3005 if isinstance(self
._entries
, list):
3008 entry
= self
._entries
[i
]
3010 entry
= self
.MissingEntry
3011 if not self
.is_incomplete
:
3012 raise self
.IndexError()
3013 if entry
is self
.MissingEntry
:
3014 raise EntryNotInPlaylist(f
'Entry {i + 1} cannot be found')
3019 return type(self
.ydl
)._handle
_extraction
_exceptions
(lambda _
, i
: self
._entries
[i
])(self
.ydl
, i
)
3020 except (LazyList
.IndexError, PagedList
.IndexError):
3021 raise self
.IndexError()
3024 def __getitem__(self
, idx
):
3025 if isinstance(idx
, int):
3026 idx
= slice(idx
, idx
)
3028 # NB: PlaylistEntries[1:10] => (0, 1, ... 9)
3029 step
= 1 if idx
.step
is None else idx
.step
3030 if idx
.start
is None:
3031 start
= 0 if step
> 0 else len(self
) - 1
3033 start
= idx
.start
- 1 if idx
.start
>= 0 else len(self
) + idx
.start
3035 # NB: Do not call len(self) when idx == [:]
3036 if idx
.stop
is None:
3037 stop
= 0 if step
< 0 else float('inf')
3039 stop
= idx
.stop
- 1 if idx
.stop
>= 0 else len(self
) + idx
.stop
3040 stop
+= [-1, 1][step
> 0]
3042 for i
in frange(start
, stop
, step
):
3046 entry
= self
._getter
(i
)
3047 except self
.IndexError:
3048 self
.is_exhausted
= True
3055 return len(tuple(self
[:]))
3057 class IndexError(IndexError):
3061 def uppercase_escape(s
):
3062 unicode_escape
= codecs
.getdecoder('unicode_escape')
3064 r
'\\U[0-9a-fA-F]{8}',
3065 lambda m
: unicode_escape(m
.group(0))[0],
3069 def lowercase_escape(s
):
3070 unicode_escape
= codecs
.getdecoder('unicode_escape')
3072 r
'\\u[0-9a-fA-F]{4}',
3073 lambda m
: unicode_escape(m
.group(0))[0],
3077 def escape_rfc3986(s
):
3078 """Escape non-ASCII characters as suggested by RFC 3986"""
3079 return urllib
.parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
3082 def escape_url(url
):
3083 """Escape URL as suggested by RFC 3986"""
3084 url_parsed
= urllib
.parse
.urlparse(url
)
3085 return url_parsed
._replace
(
3086 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
3087 path
=escape_rfc3986(url_parsed
.path
),
3088 params
=escape_rfc3986(url_parsed
.params
),
3089 query
=escape_rfc3986(url_parsed
.query
),
3090 fragment
=escape_rfc3986(url_parsed
.fragment
)
3094 def parse_qs(url
, **kwargs
):
3095 return urllib
.parse
.parse_qs(urllib
.parse
.urlparse(url
).query
, **kwargs
)
3098 def read_batch_urls(batch_fd
):
3100 if not isinstance(url
, str):
3101 url
= url
.decode('utf-8', 'replace')
3102 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
3103 for bom
in BOM_UTF8
:
3104 if url
.startswith(bom
):
3105 url
= url
[len(bom
):]
3107 if not url
or url
.startswith(('#', ';', ']')):
3109 # "#" cannot be stripped out since it is part of the URI
3110 # However, it can be safely stripped out if following a whitespace
3111 return re
.split(r
'\s#', url
, 1)[0].rstrip()
3113 with contextlib
.closing(batch_fd
) as fd
:
3114 return [url
for url
in map(fixup
, fd
) if url
]
3117 def urlencode_postdata(*args
, **kargs
):
3118 return urllib
.parse
.urlencode(*args
, **kargs
).encode('ascii')
3121 def update_url(url
, *, query_update
=None, **kwargs
):
3122 """Replace URL components specified by kwargs
3123 @param url str or parse url tuple
3124 @param query_update update query
3127 if isinstance(url
, str):
3128 if not kwargs
and not query_update
:
3131 url
= urllib
.parse
.urlparse(url
)
3133 assert 'query' not in kwargs
, 'query_update and query cannot be specified at the same time'
3134 kwargs
['query'] = urllib
.parse
.urlencode({
3135 **urllib
.parse
.parse_qs(url
.query
),
3138 return urllib
.parse
.urlunparse(url
._replace
(**kwargs
))
3141 def update_url_query(url
, query
):
3142 return update_url(url
, query_update
=query
)
3145 def update_Request(req
, url
=None, data
=None, headers
=None, query
=None):
3146 req_headers
= req
.headers
.copy()
3147 req_headers
.update(headers
or {})
3148 req_data
= data
or req
.data
3149 req_url
= update_url_query(url
or req
.get_full_url(), query
)
3150 req_get_method
= req
.get_method()
3151 if req_get_method
== 'HEAD':
3152 req_type
= HEADRequest
3153 elif req_get_method
== 'PUT':
3154 req_type
= PUTRequest
3156 req_type
= urllib
.request
.Request
3158 req_url
, data
=req_data
, headers
=req_headers
,
3159 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
3160 if hasattr(req
, 'timeout'):
3161 new_req
.timeout
= req
.timeout
3165 def _multipart_encode_impl(data
, boundary
):
3166 content_type
= 'multipart/form-data; boundary=%s' % boundary
3169 for k
, v
in data
.items():
3170 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
3171 if isinstance(k
, str):
3173 if isinstance(v
, str):
3175 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
3176 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
3177 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
3178 if boundary
.encode('ascii') in content
:
3179 raise ValueError('Boundary overlaps with data')
3182 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
3184 return out
, content_type
3187 def multipart_encode(data
, boundary
=None):
3189 Encode a dict to RFC 7578-compliant form-data
3192 A dict where keys and values can be either Unicode or bytes-like
3195 If specified a Unicode object, it's used as the boundary. Otherwise
3196 a random boundary is generated.
3198 Reference: https://tools.ietf.org/html/rfc7578
3200 has_specified_boundary
= boundary
is not None
3203 if boundary
is None:
3204 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
3207 out
, content_type
= _multipart_encode_impl(data
, boundary
)
3210 if has_specified_boundary
:
3214 return out
, content_type
3217 def is_iterable_like(x
, allowed_types
=collections
.abc
.Iterable
, blocked_types
=NO_DEFAULT
):
3218 if blocked_types
is NO_DEFAULT
:
3219 blocked_types
= (str, bytes, collections
.abc
.Mapping
)
3220 return isinstance(x
, allowed_types
) and not isinstance(x
, blocked_types
)
3223 def variadic(x
, allowed_types
=NO_DEFAULT
):
3224 return x
if is_iterable_like(x
, blocked_types
=allowed_types
) else (x
, )
3227 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
3230 val
= f(*args
, **kwargs
)
3231 except (AttributeError, KeyError, TypeError, IndexError, ValueError, ZeroDivisionError):
3234 if expected_type
is None or isinstance(val
, expected_type
):
3238 def try_get(src
, getter
, expected_type
=None):
3239 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
3242 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
3243 return {k: v for k, v in dct.items() if cndn(k, v)}
3246 def merge_dicts(*dicts
):
3248 for a_dict
in dicts
:
3249 for k
, v
in a_dict
.items():
3250 if (v
is not None and k
not in merged
3251 or isinstance(v
, str) and merged
[k
] == ''):
3256 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
3257 return string
if isinstance(string
, str) else str(string
, encoding
, errors
)
3269 TV_PARENTAL_GUIDELINES
= {
3279 def parse_age_limit(s
):
3280 # isinstance(False, int) is True. So type() must be used instead
3281 if type(s
) is int: # noqa: E721
3282 return s
if 0 <= s
<= 21 else None
3283 elif not isinstance(s
, str):
3285 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
3287 return int(m
.group('age'))
3290 return US_RATINGS
[s
]
3291 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
3293 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
3297 def strip_jsonp(code
):
3300 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
3301 (?:\s*&&\s*(?P=func_name))?
3302 \s*\(\s*(?P<callback_data>.*)\);?
3303 \s*?(?://[^\n]*)*$''',
3304 r
'\g<callback_data>', code
)
3307 def js_to_json(code
, vars={}, *, strict
=False):
3308 # vars is a dict of var, val pairs to substitute
3309 STRING_QUOTES
= '\'"`'
3310 STRING_RE
= '|'.join(rf
'{q}(?:\\.|[^\\{q}])*{q}' for q
in STRING_QUOTES
)
3311 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3312 SKIP_RE
= fr
'\s*(?:{COMMENT_RE})?\s*'
3314 (fr
'(?s)^(0[xX][0-9a-fA-F]+){SKIP_RE}:?$', 16),
3315 (fr
'(?s)^(0+[0-7]+){SKIP_RE}:?$', 8),
3318 def process_escape(match
):
3319 JSON_PASSTHROUGH_ESCAPES
= R
'"\bfnrtu'
3320 escape
= match
.group(1) or match
.group(2)
3322 return (Rf
'\{escape}' if escape
in JSON_PASSTHROUGH_ESCAPES
3323 else R
'\u00' if escape
== 'x'
3324 else '' if escape
== '\n'
3327 def template_substitute(match
):
3328 evaluated
= js_to_json(match
.group(1), vars, strict
=strict
)
3329 if evaluated
[0] == '"':
3330 return json
.loads(evaluated
)
3335 if v
in ('true', 'false', 'null'):
3337 elif v
in ('undefined', 'void 0'):
3339 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3342 if v
[0] in STRING_QUOTES
:
3343 v
= re
.sub(r
'(?s)\${([^}]+)}', template_substitute
, v
[1:-1]) if v
[0] == '`' else v
[1:-1]
3344 escaped
= re
.sub(r
'(?s)(")|\\(.)', process_escape
, v
)
3345 return f
'"{escaped}"'
3347 for regex
, base
in INTEGER_TABLE
:
3348 im
= re
.match(regex
, v
)
3350 i
= int(im
.group(1), base
)
3351 return f
'"{i}":' if v
.endswith(':') else str(i
)
3357 except json
.JSONDecodeError
:
3358 return json
.dumps(vars[v
])
3365 raise ValueError(f
'Unknown value: {v}')
3367 def create_map(mobj
):
3368 return json
.dumps(dict(json
.loads(js_to_json(mobj
.group(1) or '[]', vars=vars))))
3370 code
= re
.sub(r
'new Map\((\[.*?\])?\)', create_map
, code
)
3372 code
= re
.sub(r
'new Date\((".+")\)', r
'\g<1>', code
)
3373 code
= re
.sub(r
'new \w+\((.*?)\)', lambda m
: json
.dumps(m
.group(0)), code
)
3374 code
= re
.sub(r
'parseInt\([^\d]+(\d+)[^\d]+\)', r
'\1', code
)
3375 code
= re
.sub(r
'\(function\([^)]*\)\s*\{[^}]*\}\s*\)\s*\(\s*(["\'][^
)]*["\'])\s*\)', r'\1', code)
3377 return re.sub(rf'''(?sx)
3379 {COMMENT_RE}|,(?={SKIP_RE}[\]}}])|
3380 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3381 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{SKIP_RE}:)?|
3382 [0-9]+(?={SKIP_RE}:)|
3387 def qualities(quality_ids):
3388 """ Get a numeric quality value out of a list of possible values """
3391 return quality_ids.index(qid)
3397 POSTPROCESS_WHEN = ('pre_process', 'after_filter', 'video', 'before_dl', 'post_process', 'after_move', 'after_video', 'playlist')
3401 'default': '%(title)s [%(id)s].%(ext)s',
3402 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3408 'description': 'description',
3409 'annotation': 'annotations.xml',
3410 'infojson': 'info.json',
3413 'pl_thumbnail': None,
3414 'pl_description': 'description',
3415 'pl_infojson': 'info.json',
3418 # As of [1] format syntax is:
3419 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3420 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3421 STR_FORMAT_RE_TMPL = r'''(?x)
3422 (?<!%)(?P<prefix>(?:%%)*)
3424 (?P<has_key>\((?P<key>{0})\))?
3426 (?P<conversion>[#0\-+ ]+)?
3428 (?P<precision>\.\d+)?
3429 (?P<len_mod>[hlL])? # unused in python
3430 {1} # conversion type
3435 STR_FORMAT_TYPES = 'diouxXeEfFgGcrs'
3438 def limit_length(s, length):
3439 """ Add ellipses to overly long strings """
3444 return s[:length - len(ELLIPSES)] + ELLIPSES
3448 def version_tuple(v):
3449 return tuple(int(e) for e in re.split(r'[-.]', v))
3452 def is_outdated_version(version, limit, assume_new=True):
3454 return not assume_new
3456 return version_tuple(version) < version_tuple(limit)
3458 return not assume_new
3461 def ytdl_is_updateable():
3462 """ Returns if yt-dlp can be updated with -U """
3464 from ..update import is_non_updateable
3466 return not is_non_updateable()
3469 def args_to_str(args):
3470 # Get a short string representation for a subprocess command
3471 return ' '.join(compat_shlex_quote(a) for a in args)
3474 def error_to_str(err):
3475 return f'{type(err).__name__}: {err}'
3478 def mimetype2ext(mt, default=NO_DEFAULT):
3479 if not isinstance(mt, str):
3480 if default is not NO_DEFAULT:
3496 'x-matroska': 'mkv',
3498 'x-mp4-fragmented': 'mp4',
3503 # application (streaming playlists)
3507 'vnd.apple.mpegurl': 'm3u8',
3508 'vnd.ms-sstr+xml': 'ism',
3509 'x-mpegurl': 'm3u8',
3513 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3.
3514 # Using .mp3 as it's the most popular one
3515 'audio/mpeg': 'mp3',
3516 'audio/webm': 'webm',
3517 'audio/x-matroska': 'mka',
3518 'audio/x-mpegurl': 'm3u',
3526 'x-realaudio': 'ra',
3537 'vnd.wap.wbmp': 'wbmp',
3544 'filmstrip+json': 'fs',
3545 'smptett+xml': 'tt',
3548 'x-ms-sami': 'sami',
3557 mimetype = mt.partition(';')[0].strip().lower()
3558 _, _, subtype = mimetype.rpartition('/')
3560 ext = traversal.traverse_obj(MAP, mimetype, subtype, subtype.rsplit('+')[-1])
3563 elif default is not NO_DEFAULT:
3565 return subtype.replace('+', '.')
3568 def ext2mimetype(ext_or_url):
3571 if '.' not in ext_or_url:
3572 ext_or_url = f'file.{ext_or_url}'
3573 return mimetypes.guess_type(ext_or_url)[0]
3576 def parse_codecs(codecs_str):
3577 # http://tools.ietf.org/html/rfc6381
3580 split_codecs = list(filter(None, map(
3581 str.strip, codecs_str.strip().strip(',').split(','))))
3582 vcodec, acodec, scodec, hdr = None, None, None, None
3583 for full_codec in split_codecs:
3584 parts = re.sub(r'0+(?=\d)', '', full_codec).split('.')
3585 if parts[0] in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3586 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3590 if parts[0] in ('dvh1', 'dvhe'):
3592 elif parts[0] == 'av1' and traversal.traverse_obj(parts, 3) == '10':
3594 elif parts[:2] == ['vp9', '2']:
3596 elif parts[0] in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-4',
3597 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3598 acodec = acodec or full_codec
3599 elif parts[0] in ('stpp', 'wvtt'):
3600 scodec = scodec or full_codec
3602 write_string(f'WARNING: Unknown codec {full_codec}\n')
3603 if vcodec or acodec or scodec:
3605 'vcodec': vcodec or 'none',
3606 'acodec': acodec or 'none',
3607 'dynamic_range': hdr,
3608 **({'scodec': scodec} if scodec is not None else {}),
3610 elif len(split_codecs) == 2:
3612 'vcodec': split_codecs[0],
3613 'acodec': split_codecs[1],
3618 def get_compatible_ext(*, vcodecs, acodecs, vexts, aexts, preferences=None):
3619 assert len(vcodecs) == len(vexts) and len(acodecs) == len(aexts)
3621 allow_mkv = not preferences or 'mkv' in preferences
3623 if allow_mkv and max(len(acodecs), len(vcodecs)) > 1:
3624 return 'mkv' # TODO: any other format allows this?
3626 # TODO: All codecs supported by parse_codecs isn't handled here
3627 COMPATIBLE_CODECS = {
3629 'av1', 'hevc', 'avc1', 'mp4a', 'ac-4', # fourcc (m3u8, mpd)
3630 'h264', 'aacl', 'ec-3', # Set in ISM
3633 'av1', 'vp9', 'vp8', 'opus', 'vrbs',
3634 'vp9x', 'vp8x', # in the webm spec
3638 sanitize_codec = functools.partial(try_get, getter=lambda x: x[0].split('.')[0].replace('0', ''))
3639 vcodec, acodec = sanitize_codec(vcodecs), sanitize_codec(acodecs)
3641 for ext in preferences or COMPATIBLE_CODECS.keys():
3642 codec_set = COMPATIBLE_CODECS.get(ext, set())
3643 if ext == 'mkv' or codec_set.issuperset((vcodec, acodec)):
3647 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma', 'mov'},
3650 for ext in preferences or vexts:
3651 current_exts = {ext, *vexts, *aexts}
3652 if ext == 'mkv' or current_exts == {ext} or any(
3653 ext_sets.issuperset(current_exts) for ext_sets in COMPATIBLE_EXTS):
3655 return 'mkv' if allow_mkv else preferences[-1]
3658 def urlhandle_detect_ext(url_handle, default=NO_DEFAULT):
3659 getheader = url_handle.headers.get
3661 cd = getheader('Content-Disposition')
3663 m = re.match(r'attachment;\s*filename="(?P
<filename
>[^
"]+)"', cd)
3665 e = determine_ext(m.group('filename
'), default_ext=None)
3669 meta_ext = getheader('x
-amz
-meta
-name
')
3671 e = meta_ext.rpartition('.')[2]
3675 return mimetype2ext(getheader('Content
-Type
'), default=default)
3678 def encode_data_uri(data, mime_type):
3679 return 'data
:%s;base64
,%s' % (mime_type, base64.b64encode(data).decode('ascii
'))
3682 def age_restricted(content_limit, age_limit):
3683 """ Returns True iff the content should be blocked """
3685 if age_limit is None: # No limit set
3687 if content_limit is None:
3688 return False # Content available for everyone
3689 return age_limit < content_limit
3692 # List of known byte-order-marks (BOM)
3694 (b'\xef\xbb\xbf', 'utf
-8'),
3695 (b'\x00\x00\xfe\xff', 'utf
-32-be
'),
3696 (b'\xff\xfe\x00\x00', 'utf
-32-le
'),
3697 (b'\xff\xfe', 'utf
-16-le
'),
3698 (b'\xfe\xff', 'utf
-16-be
'),
3702 def is_html(first_bytes):
3703 """ Detect whether a file contains HTML by examining its first bytes. """
3706 for bom, enc in BOMS:
3707 while first_bytes.startswith(bom):
3708 encoding, first_bytes = enc, first_bytes[len(bom):]
3710 return re.match(r'^\s
*<', first_bytes.decode(encoding, 'replace
'))
3713 def determine_protocol(info_dict):
3714 protocol = info_dict.get('protocol
')
3715 if protocol is not None:
3718 url = sanitize_url(info_dict['url
'])
3719 if url.startswith('rtmp
'):
3721 elif url.startswith('mms
'):
3723 elif url.startswith('rtsp
'):
3726 ext = determine_ext(url)
3728 return 'm3u8
' if info_dict.get('is_live
') else 'm3u8_native
'
3732 return urllib.parse.urlparse(url).scheme
3735 def render_table(header_row, data, delim=False, extra_gap=0, hide_empty=False):
3736 """ Render a list of rows, each as a list of values.
3737 Text after a \t will be right aligned """
3739 return len(remove_terminal_sequences(string).replace('\t', ''))
3741 def get_max_lens(table):
3742 return [max(width(str(v)) for v in col) for col in zip(*table)]
3744 def filter_using_list(row, filterArray):
3745 return [col for take, col in itertools.zip_longest(filterArray, row, fillvalue=True) if take]
3747 max_lens = get_max_lens(data) if hide_empty else []
3748 header_row = filter_using_list(header_row, max_lens)
3749 data = [filter_using_list(row, max_lens) for row in data]
3751 table = [header_row] + data
3752 max_lens = get_max_lens(table)
3755 table = [header_row, [delim * (ml + extra_gap) for ml in max_lens]] + data
3756 table[1][-1] = table[1][-1][:-extra_gap * len(delim)] # Remove extra_gap from end of delimiter
3758 for pos, text in enumerate(map(str, row)):
3760 row[pos] = text.replace('\t', ' ' * (max_lens[pos] - width(text))) + ' ' * extra_gap
3762 row[pos] = text + ' ' * (max_lens[pos] - width(text) + extra_gap)
3763 ret = '\n'.join(''.join(row).rstrip() for row in table)
3767 def _match_one(filter_part, dct, incomplete):
3768 # TODO: Generalize code with YoutubeDL._build_format_filter
3769 STRING_OPERATORS = {
3770 '*=': operator.contains,
3771 '^
=': lambda attr, value: attr.startswith(value),
3772 '$
=': lambda attr, value: attr.endswith(value),
3773 '~
=': lambda attr, value: re.search(value, attr),
3775 COMPARISON_OPERATORS = {
3777 '<=': operator.le, # "<=" must be defined above "<"
3784 if isinstance(incomplete, bool):
3785 is_incomplete = lambda _: incomplete
3787 is_incomplete = lambda k: k in incomplete
3789 operator_rex = re.compile(r'''(?x)
3791 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3793 (?P<quote>["\'])(?P<quotedstrval>.+?)(?P=quote)|
3796 ''' % '|
'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3797 m = operator_rex.fullmatch(filter_part.strip())
3800 unnegated_op = COMPARISON_OPERATORS[m['op
']]
3802 op = lambda attr, value: not unnegated_op(attr, value)
3805 comparison_value = m['quotedstrval
'] or m['strval
'] or m['intval
']
3807 comparison_value = comparison_value.replace(r'\
%s' % m['quote
'], m['quote
'])
3808 actual_value = dct.get(m['key
'])
3809 numeric_comparison = None
3810 if isinstance(actual_value, (int, float)):
3811 # If the original field is a string and matching comparisonvalue is
3812 # a number we should respect the origin of the original field
3813 # and process comparison value as a string (see
3814 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3816 numeric_comparison = int(comparison_value)
3818 numeric_comparison = parse_filesize(comparison_value)
3819 if numeric_comparison is None:
3820 numeric_comparison = parse_filesize(f'{comparison_value}B
')
3821 if numeric_comparison is None:
3822 numeric_comparison = parse_duration(comparison_value)
3823 if numeric_comparison is not None and m['op
'] in STRING_OPERATORS:
3824 raise ValueError('Operator
%s only supports string values
!' % m['op
'])
3825 if actual_value is None:
3826 return is_incomplete(m['key
']) or m['none_inclusive
']
3827 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3830 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3831 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3833 operator_rex = re.compile(r'''(?x)
3834 (?P<op>%s)\s*(?P<key>[a-z_]+)
3835 ''' % '|
'.join(map(re.escape, UNARY_OPERATORS.keys())))
3836 m = operator_rex.fullmatch(filter_part.strip())
3838 op = UNARY_OPERATORS[m.group('op
')]
3839 actual_value = dct.get(m.group('key
'))
3840 if is_incomplete(m.group('key
')) and actual_value is None:
3842 return op(actual_value)
3844 raise ValueError('Invalid
filter part
%r' % filter_part)
3847 def match_str(filter_str, dct, incomplete=False):
3848 """ Filter a dictionary with a simple string syntax.
3849 @returns Whether the filter passes
3850 @param incomplete Set of keys that is expected to be missing from dct.
3851 Can be True/False to indicate all/none of the keys may be missing.
3852 All conditions on incomplete keys pass if the key is missing
3855 _match_one(filter_part.replace(r'\
&', '&'), dct, incomplete)
3856 for filter_part in re.split(r'(?
<!\\)&', filter_str))
3859 def match_filter_func(filters, breaking_filters=None):
3860 if not filters and not breaking_filters:
3862 breaking_filters = match_filter_func(breaking_filters) or (lambda _, __: None)
3863 filters = set(variadic(filters or []))
3865 interactive = '-' in filters
3869 def _match_func(info_dict, incomplete=False):
3870 ret = breaking_filters(info_dict, incomplete)
3872 raise RejectedVideoReached(ret)
3874 if not filters or any(match_str(f, info_dict, incomplete) for f in filters):
3875 return NO_DEFAULT if interactive and not incomplete else None
3877 video_title = info_dict.get('title
') or info_dict.get('id') or 'entry
'
3878 filter_str = ') |
('.join(map(str.strip, filters))
3879 return f'{video_title} does
not pass filter ({filter_str}
), skipping
..'
3883 class download_range_func:
3884 def __init__(self, chapters, ranges):
3885 self.chapters, self.ranges = chapters, ranges
3887 def __call__(self, info_dict, ydl):
3888 if not self.ranges and not self.chapters:
3891 warning = ('There are no chapters matching the regex
' if info_dict.get('chapters
')
3892 else 'Cannot match chapters since chapter information
is unavailable
')
3893 for regex in self.chapters or []:
3894 for i, chapter in enumerate(info_dict.get('chapters
') or []):
3895 if re.search(regex, chapter['title
']):
3897 yield {**chapter, 'index': i}
3898 if self.chapters and warning:
3899 ydl.to_screen(f'[info
] {info_dict["id"]}
: {warning}
')
3901 yield from ({'start_time': start, 'end_time': end} for start, end in self.ranges or [])
3903 def __eq__(self, other):
3904 return (isinstance(other, download_range_func)
3905 and self.chapters == other.chapters and self.ranges == other.ranges)
3908 return f'{__name__}
.{type(self).__name__}
({self.chapters}
, {self.ranges}
)'
3911 def parse_dfxp_time_expr(time_expr):
3915 mobj = re.match(rf'^
(?P
<time_offset
>{NUMBER_RE}
)s?$
', time_expr)
3917 return float(mobj.group('time_offset
'))
3919 mobj = re.match(r'^
(\d
+):(\d\d
):(\d\
d(?
:(?
:\
.|
:)\d
+)?
)$
', time_expr)
3921 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3924 def srt_subtitles_timecode(seconds):
3925 return '%02d
:%02d
:%02d
,%03d
' % timetuple_from_msec(seconds * 1000)
3928 def ass_subtitles_timecode(seconds):
3929 time = timetuple_from_msec(seconds * 1000)
3930 return '%01d
:%02d
:%02d
.%02d
' % (*time[:-1], time.milliseconds / 10)
3933 def dfxp2srt(dfxp_data):
3935 @param dfxp_data A bytes-like object containing DFXP data
3936 @returns A unicode object containing converted SRT data
3938 LEGACY_NAMESPACES = (
3939 (b'http
://www
.w3
.org
/ns
/ttml
', [
3940 b'http
://www
.w3
.org
/2004/11/ttaf1
',
3941 b'http
://www
.w3
.org
/2006/04/ttaf1
',
3942 b'http
://www
.w3
.org
/2006/10/ttaf1
',
3944 (b'http
://www
.w3
.org
/ns
/ttml
#styling', [
3945 b
'http://www.w3.org/ns/ttml#style',
3949 SUPPORTED_STYLING
= [
3958 _x
= functools
.partial(xpath_with_ns
, ns_map
={
3959 'xml': 'http://www.w3.org/XML/1998/namespace',
3960 'ttml': 'http://www.w3.org/ns/ttml',
3961 'tts': 'http://www.w3.org/ns/ttml#styling',
3967 class TTMLPElementParser
:
3969 _unclosed_elements
= []
3970 _applied_styles
= []
3972 def start(self
, tag
, attrib
):
3973 if tag
in (_x('ttml:br'), 'br'):
3976 unclosed_elements
= []
3978 element_style_id
= attrib
.get('style')
3980 style
.update(default_style
)
3981 if element_style_id
:
3982 style
.update(styles
.get(element_style_id
, {}))
3983 for prop
in SUPPORTED_STYLING
:
3984 prop_val
= attrib
.get(_x('tts:' + prop
))
3986 style
[prop
] = prop_val
3989 for k
, v
in sorted(style
.items()):
3990 if self
._applied
_styles
and self
._applied
_styles
[-1].get(k
) == v
:
3993 font
+= ' color="%s"' % v
3994 elif k
== 'fontSize':
3995 font
+= ' size="%s"' % v
3996 elif k
== 'fontFamily':
3997 font
+= ' face="%s"' % v
3998 elif k
== 'fontWeight' and v
== 'bold':
4000 unclosed_elements
.append('b')
4001 elif k
== 'fontStyle' and v
== 'italic':
4003 unclosed_elements
.append('i')
4004 elif k
== 'textDecoration' and v
== 'underline':
4006 unclosed_elements
.append('u')
4008 self
._out
+= '<font' + font
+ '>'
4009 unclosed_elements
.append('font')
4011 if self
._applied
_styles
:
4012 applied_style
.update(self
._applied
_styles
[-1])
4013 applied_style
.update(style
)
4014 self
._applied
_styles
.append(applied_style
)
4015 self
._unclosed
_elements
.append(unclosed_elements
)
4018 if tag
not in (_x('ttml:br'), 'br'):
4019 unclosed_elements
= self
._unclosed
_elements
.pop()
4020 for element
in reversed(unclosed_elements
):
4021 self
._out
+= '</%s>' % element
4022 if unclosed_elements
and self
._applied
_styles
:
4023 self
._applied
_styles
.pop()
4025 def data(self
, data
):
4029 return self
._out
.strip()
4031 # Fix UTF-8 encoded file wrongly marked as UTF-16. See https://github.com/yt-dlp/yt-dlp/issues/6543#issuecomment-1477169870
4032 # This will not trigger false positives since only UTF-8 text is being replaced
4033 dfxp_data
= dfxp_data
.replace(b
'encoding=\'UTF-16\'', b
'encoding=\'UTF-8\'')
4035 def parse_node(node
):
4036 target
= TTMLPElementParser()
4037 parser
= xml
.etree
.ElementTree
.XMLParser(target
=target
)
4038 parser
.feed(xml
.etree
.ElementTree
.tostring(node
))
4039 return parser
.close()
4041 for k
, v
in LEGACY_NAMESPACES
:
4043 dfxp_data
= dfxp_data
.replace(ns
, k
)
4045 dfxp
= compat_etree_fromstring(dfxp_data
)
4047 paras
= dfxp
.findall(_x('.//ttml:p')) or dfxp
.findall('.//p')
4050 raise ValueError('Invalid dfxp/TTML subtitle')
4054 for style
in dfxp
.findall(_x('.//ttml:style')):
4055 style_id
= style
.get('id') or style
.get(_x('xml:id'))
4058 parent_style_id
= style
.get('style')
4060 if parent_style_id
not in styles
:
4063 styles
[style_id
] = styles
[parent_style_id
].copy()
4064 for prop
in SUPPORTED_STYLING
:
4065 prop_val
= style
.get(_x('tts:' + prop
))
4067 styles
.setdefault(style_id
, {})[prop
] = prop_val
4073 for p
in ('body', 'div'):
4074 ele
= xpath_element(dfxp
, [_x('.//ttml:' + p
), './/' + p
])
4077 style
= styles
.get(ele
.get('style'))
4080 default_style
.update(style
)
4082 for para
, index
in zip(paras
, itertools
.count(1)):
4083 begin_time
= parse_dfxp_time_expr(para
.attrib
.get('begin'))
4084 end_time
= parse_dfxp_time_expr(para
.attrib
.get('end'))
4085 dur
= parse_dfxp_time_expr(para
.attrib
.get('dur'))
4086 if begin_time
is None:
4091 end_time
= begin_time
+ dur
4092 out
.append('%d\n%s --> %s\n%s\n\n' % (
4094 srt_subtitles_timecode(begin_time
),
4095 srt_subtitles_timecode(end_time
),
4101 def cli_option(params
, command_option
, param
, separator
=None):
4102 param
= params
.get(param
)
4103 return ([] if param
is None
4104 else [command_option
, str(param
)] if separator
is None
4105 else [f
'{command_option}{separator}{param}'])
4108 def cli_bool_option(params
, command_option
, param
, true_value
='true', false_value
='false', separator
=None):
4109 param
= params
.get(param
)
4110 assert param
in (True, False, None)
4111 return cli_option({True: true_value, False: false_value}
, command_option
, param
, separator
)
4114 def cli_valueless_option(params
, command_option
, param
, expected_value
=True):
4115 return [command_option
] if params
.get(param
) == expected_value
else []
4118 def cli_configuration_args(argdict
, keys
, default
=[], use_compat
=True):
4119 if isinstance(argdict
, (list, tuple)): # for backward compatibility
4126 assert isinstance(argdict
, dict)
4128 assert isinstance(keys
, (list, tuple))
4129 for key_list
in keys
:
4130 arg_list
= list(filter(
4131 lambda x
: x
is not None,
4132 [argdict
.get(key
.lower()) for key
in variadic(key_list
)]))
4134 return [arg
for args
in arg_list
for arg
in args
]
4138 def _configuration_args(main_key
, argdict
, exe
, keys
=None, default
=[], use_compat
=True):
4139 main_key
, exe
= main_key
.lower(), exe
.lower()
4140 root_key
= exe
if main_key
== exe
else f
'{main_key}+{exe}'
4141 keys
= [f
'{root_key}{k}' for k
in (keys
or [''])]
4142 if root_key
in keys
:
4144 keys
.append((main_key
, exe
))
4145 keys
.append('default')
4148 return cli_configuration_args(argdict
, keys
, default
, use_compat
)
4152 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
4211 'iw': 'heb', # Replaced by he in 1989 revision
4221 'in': 'ind', # Replaced by id in 1989 revision
4336 'ji': 'yid', # Replaced by yi in 1989 revision
4344 def short2long(cls
, code
):
4345 """Convert language code from ISO 639-1 to ISO 639-2/T"""
4346 return cls
._lang
_map
.get(code
[:2])
4349 def long2short(cls
, code
):
4350 """Convert language code from ISO 639-2/T to ISO 639-1"""
4351 for short_name
, long_name
in cls
._lang
_map
.items():
4352 if long_name
== code
:
4357 # From http://data.okfn.org/data/core/country-list
4359 'AF': 'Afghanistan',
4360 'AX': 'Åland Islands',
4363 'AS': 'American Samoa',
4368 'AG': 'Antigua and Barbuda',
4385 'BO': 'Bolivia, Plurinational State of',
4386 'BQ': 'Bonaire, Sint Eustatius and Saba',
4387 'BA': 'Bosnia and Herzegovina',
4389 'BV': 'Bouvet Island',
4391 'IO': 'British Indian Ocean Territory',
4392 'BN': 'Brunei Darussalam',
4394 'BF': 'Burkina Faso',
4400 'KY': 'Cayman Islands',
4401 'CF': 'Central African Republic',
4405 'CX': 'Christmas Island',
4406 'CC': 'Cocos (Keeling) Islands',
4410 'CD': 'Congo, the Democratic Republic of the',
4411 'CK': 'Cook Islands',
4413 'CI': 'Côte d\'Ivoire',
4418 'CZ': 'Czech Republic',
4422 'DO': 'Dominican Republic',
4425 'SV': 'El Salvador',
4426 'GQ': 'Equatorial Guinea',
4430 'FK': 'Falkland Islands (Malvinas)',
4431 'FO': 'Faroe Islands',
4435 'GF': 'French Guiana',
4436 'PF': 'French Polynesia',
4437 'TF': 'French Southern Territories',
4452 'GW': 'Guinea-Bissau',
4455 'HM': 'Heard Island and McDonald Islands',
4456 'VA': 'Holy See (Vatican City State)',
4463 'IR': 'Iran, Islamic Republic of',
4466 'IM': 'Isle of Man',
4476 'KP': 'Korea, Democratic People\'s Republic of',
4477 'KR': 'Korea, Republic of',
4480 'LA': 'Lao People\'s Democratic Republic',
4486 'LI': 'Liechtenstein',
4490 'MK': 'Macedonia, the Former Yugoslav Republic of',
4497 'MH': 'Marshall Islands',
4503 'FM': 'Micronesia, Federated States of',
4504 'MD': 'Moldova, Republic of',
4515 'NL': 'Netherlands',
4516 'NC': 'New Caledonia',
4517 'NZ': 'New Zealand',
4522 'NF': 'Norfolk Island',
4523 'MP': 'Northern Mariana Islands',
4528 'PS': 'Palestine, State of',
4530 'PG': 'Papua New Guinea',
4533 'PH': 'Philippines',
4537 'PR': 'Puerto Rico',
4541 'RU': 'Russian Federation',
4543 'BL': 'Saint Barthélemy',
4544 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4545 'KN': 'Saint Kitts and Nevis',
4546 'LC': 'Saint Lucia',
4547 'MF': 'Saint Martin (French part)',
4548 'PM': 'Saint Pierre and Miquelon',
4549 'VC': 'Saint Vincent and the Grenadines',
4552 'ST': 'Sao Tome and Principe',
4553 'SA': 'Saudi Arabia',
4557 'SL': 'Sierra Leone',
4559 'SX': 'Sint Maarten (Dutch part)',
4562 'SB': 'Solomon Islands',
4564 'ZA': 'South Africa',
4565 'GS': 'South Georgia and the South Sandwich Islands',
4566 'SS': 'South Sudan',
4571 'SJ': 'Svalbard and Jan Mayen',
4574 'CH': 'Switzerland',
4575 'SY': 'Syrian Arab Republic',
4576 'TW': 'Taiwan, Province of China',
4578 'TZ': 'Tanzania, United Republic of',
4580 'TL': 'Timor-Leste',
4584 'TT': 'Trinidad and Tobago',
4587 'TM': 'Turkmenistan',
4588 'TC': 'Turks and Caicos Islands',
4592 'AE': 'United Arab Emirates',
4593 'GB': 'United Kingdom',
4594 'US': 'United States',
4595 'UM': 'United States Minor Outlying Islands',
4599 'VE': 'Venezuela, Bolivarian Republic of',
4601 'VG': 'Virgin Islands, British',
4602 'VI': 'Virgin Islands, U.S.',
4603 'WF': 'Wallis and Futuna',
4604 'EH': 'Western Sahara',
4608 # Not ISO 3166 codes, but used for IP blocks
4609 'AP': 'Asia/Pacific Region',
4614 def short2full(cls
, code
):
4615 """Convert an ISO 3166-2 country code to the corresponding full name"""
4616 return cls
._country
_map
.get(code
.upper())
4620 # Major IPv4 address blocks per country
4622 'AD': '46.172.224.0/19',
4623 'AE': '94.200.0.0/13',
4624 'AF': '149.54.0.0/17',
4625 'AG': '209.59.64.0/18',
4626 'AI': '204.14.248.0/21',
4627 'AL': '46.99.0.0/16',
4628 'AM': '46.70.0.0/15',
4629 'AO': '105.168.0.0/13',
4630 'AP': '182.50.184.0/21',
4631 'AQ': '23.154.160.0/24',
4632 'AR': '181.0.0.0/12',
4633 'AS': '202.70.112.0/20',
4634 'AT': '77.116.0.0/14',
4635 'AU': '1.128.0.0/11',
4636 'AW': '181.41.0.0/18',
4637 'AX': '185.217.4.0/22',
4638 'AZ': '5.197.0.0/16',
4639 'BA': '31.176.128.0/17',
4640 'BB': '65.48.128.0/17',
4641 'BD': '114.130.0.0/16',
4643 'BF': '102.178.0.0/15',
4644 'BG': '95.42.0.0/15',
4645 'BH': '37.131.0.0/17',
4646 'BI': '154.117.192.0/18',
4647 'BJ': '137.255.0.0/16',
4648 'BL': '185.212.72.0/23',
4649 'BM': '196.12.64.0/18',
4650 'BN': '156.31.0.0/16',
4651 'BO': '161.56.0.0/16',
4652 'BQ': '161.0.80.0/20',
4653 'BR': '191.128.0.0/12',
4654 'BS': '24.51.64.0/18',
4655 'BT': '119.2.96.0/19',
4656 'BW': '168.167.0.0/16',
4657 'BY': '178.120.0.0/13',
4658 'BZ': '179.42.192.0/18',
4659 'CA': '99.224.0.0/11',
4660 'CD': '41.243.0.0/16',
4661 'CF': '197.242.176.0/21',
4662 'CG': '160.113.0.0/16',
4663 'CH': '85.0.0.0/13',
4664 'CI': '102.136.0.0/14',
4665 'CK': '202.65.32.0/19',
4666 'CL': '152.172.0.0/14',
4667 'CM': '102.244.0.0/14',
4668 'CN': '36.128.0.0/10',
4669 'CO': '181.240.0.0/12',
4670 'CR': '201.192.0.0/12',
4671 'CU': '152.206.0.0/15',
4672 'CV': '165.90.96.0/19',
4673 'CW': '190.88.128.0/17',
4674 'CY': '31.153.0.0/16',
4675 'CZ': '88.100.0.0/14',
4677 'DJ': '197.241.0.0/17',
4678 'DK': '87.48.0.0/12',
4679 'DM': '192.243.48.0/20',
4680 'DO': '152.166.0.0/15',
4681 'DZ': '41.96.0.0/12',
4682 'EC': '186.68.0.0/15',
4683 'EE': '90.190.0.0/15',
4684 'EG': '156.160.0.0/11',
4685 'ER': '196.200.96.0/20',
4686 'ES': '88.0.0.0/11',
4687 'ET': '196.188.0.0/14',
4688 'EU': '2.16.0.0/13',
4689 'FI': '91.152.0.0/13',
4690 'FJ': '144.120.0.0/16',
4691 'FK': '80.73.208.0/21',
4692 'FM': '119.252.112.0/20',
4693 'FO': '88.85.32.0/19',
4695 'GA': '41.158.0.0/15',
4697 'GD': '74.122.88.0/21',
4698 'GE': '31.146.0.0/16',
4699 'GF': '161.22.64.0/18',
4700 'GG': '62.68.160.0/19',
4701 'GH': '154.160.0.0/12',
4702 'GI': '95.164.0.0/16',
4703 'GL': '88.83.0.0/19',
4704 'GM': '160.182.0.0/15',
4705 'GN': '197.149.192.0/18',
4706 'GP': '104.250.0.0/19',
4707 'GQ': '105.235.224.0/20',
4708 'GR': '94.64.0.0/13',
4709 'GT': '168.234.0.0/16',
4710 'GU': '168.123.0.0/16',
4711 'GW': '197.214.80.0/20',
4712 'GY': '181.41.64.0/18',
4713 'HK': '113.252.0.0/14',
4714 'HN': '181.210.0.0/16',
4715 'HR': '93.136.0.0/13',
4716 'HT': '148.102.128.0/17',
4717 'HU': '84.0.0.0/14',
4718 'ID': '39.192.0.0/10',
4719 'IE': '87.32.0.0/12',
4720 'IL': '79.176.0.0/13',
4721 'IM': '5.62.80.0/20',
4722 'IN': '117.192.0.0/10',
4723 'IO': '203.83.48.0/21',
4724 'IQ': '37.236.0.0/14',
4725 'IR': '2.176.0.0/12',
4726 'IS': '82.221.0.0/16',
4727 'IT': '79.0.0.0/10',
4728 'JE': '87.244.64.0/18',
4729 'JM': '72.27.0.0/17',
4730 'JO': '176.29.0.0/16',
4731 'JP': '133.0.0.0/8',
4732 'KE': '105.48.0.0/12',
4733 'KG': '158.181.128.0/17',
4734 'KH': '36.37.128.0/17',
4735 'KI': '103.25.140.0/22',
4736 'KM': '197.255.224.0/20',
4737 'KN': '198.167.192.0/19',
4738 'KP': '175.45.176.0/22',
4739 'KR': '175.192.0.0/10',
4740 'KW': '37.36.0.0/14',
4741 'KY': '64.96.0.0/15',
4742 'KZ': '2.72.0.0/13',
4743 'LA': '115.84.64.0/18',
4744 'LB': '178.135.0.0/16',
4745 'LC': '24.92.144.0/20',
4746 'LI': '82.117.0.0/19',
4747 'LK': '112.134.0.0/15',
4748 'LR': '102.183.0.0/16',
4749 'LS': '129.232.0.0/17',
4750 'LT': '78.56.0.0/13',
4751 'LU': '188.42.0.0/16',
4752 'LV': '46.109.0.0/16',
4753 'LY': '41.252.0.0/14',
4754 'MA': '105.128.0.0/11',
4755 'MC': '88.209.64.0/18',
4756 'MD': '37.246.0.0/16',
4757 'ME': '178.175.0.0/17',
4758 'MF': '74.112.232.0/21',
4759 'MG': '154.126.0.0/17',
4760 'MH': '117.103.88.0/21',
4761 'MK': '77.28.0.0/15',
4762 'ML': '154.118.128.0/18',
4763 'MM': '37.111.0.0/17',
4764 'MN': '49.0.128.0/17',
4765 'MO': '60.246.0.0/16',
4766 'MP': '202.88.64.0/20',
4767 'MQ': '109.203.224.0/19',
4768 'MR': '41.188.64.0/18',
4769 'MS': '208.90.112.0/22',
4770 'MT': '46.11.0.0/16',
4771 'MU': '105.16.0.0/12',
4772 'MV': '27.114.128.0/18',
4773 'MW': '102.70.0.0/15',
4774 'MX': '187.192.0.0/11',
4775 'MY': '175.136.0.0/13',
4776 'MZ': '197.218.0.0/15',
4777 'NA': '41.182.0.0/16',
4778 'NC': '101.101.0.0/18',
4779 'NE': '197.214.0.0/18',
4780 'NF': '203.17.240.0/22',
4781 'NG': '105.112.0.0/12',
4782 'NI': '186.76.0.0/15',
4783 'NL': '145.96.0.0/11',
4784 'NO': '84.208.0.0/13',
4785 'NP': '36.252.0.0/15',
4786 'NR': '203.98.224.0/19',
4787 'NU': '49.156.48.0/22',
4788 'NZ': '49.224.0.0/14',
4789 'OM': '5.36.0.0/15',
4790 'PA': '186.72.0.0/15',
4791 'PE': '186.160.0.0/14',
4792 'PF': '123.50.64.0/18',
4793 'PG': '124.240.192.0/19',
4794 'PH': '49.144.0.0/13',
4795 'PK': '39.32.0.0/11',
4796 'PL': '83.0.0.0/11',
4797 'PM': '70.36.0.0/20',
4798 'PR': '66.50.0.0/16',
4799 'PS': '188.161.0.0/16',
4800 'PT': '85.240.0.0/13',
4801 'PW': '202.124.224.0/20',
4802 'PY': '181.120.0.0/14',
4803 'QA': '37.210.0.0/15',
4804 'RE': '102.35.0.0/16',
4805 'RO': '79.112.0.0/13',
4806 'RS': '93.86.0.0/15',
4807 'RU': '5.136.0.0/13',
4808 'RW': '41.186.0.0/16',
4809 'SA': '188.48.0.0/13',
4810 'SB': '202.1.160.0/19',
4811 'SC': '154.192.0.0/11',
4812 'SD': '102.120.0.0/13',
4813 'SE': '78.64.0.0/12',
4814 'SG': '8.128.0.0/10',
4815 'SI': '188.196.0.0/14',
4816 'SK': '78.98.0.0/15',
4817 'SL': '102.143.0.0/17',
4818 'SM': '89.186.32.0/19',
4819 'SN': '41.82.0.0/15',
4820 'SO': '154.115.192.0/18',
4821 'SR': '186.179.128.0/17',
4822 'SS': '105.235.208.0/21',
4823 'ST': '197.159.160.0/19',
4824 'SV': '168.243.0.0/16',
4825 'SX': '190.102.0.0/20',
4827 'SZ': '41.84.224.0/19',
4828 'TC': '65.255.48.0/20',
4829 'TD': '154.68.128.0/19',
4830 'TG': '196.168.0.0/14',
4831 'TH': '171.96.0.0/13',
4832 'TJ': '85.9.128.0/18',
4833 'TK': '27.96.24.0/21',
4834 'TL': '180.189.160.0/20',
4835 'TM': '95.85.96.0/19',
4836 'TN': '197.0.0.0/11',
4837 'TO': '175.176.144.0/21',
4838 'TR': '78.160.0.0/11',
4839 'TT': '186.44.0.0/15',
4840 'TV': '202.2.96.0/19',
4841 'TW': '120.96.0.0/11',
4842 'TZ': '156.156.0.0/14',
4843 'UA': '37.52.0.0/14',
4844 'UG': '102.80.0.0/13',
4846 'UY': '167.56.0.0/13',
4847 'UZ': '84.54.64.0/18',
4848 'VA': '212.77.0.0/19',
4849 'VC': '207.191.240.0/21',
4850 'VE': '186.88.0.0/13',
4851 'VG': '66.81.192.0/20',
4852 'VI': '146.226.0.0/16',
4853 'VN': '14.160.0.0/11',
4854 'VU': '202.80.32.0/20',
4855 'WF': '117.20.32.0/21',
4856 'WS': '202.4.32.0/19',
4857 'YE': '134.35.0.0/16',
4858 'YT': '41.242.116.0/22',
4859 'ZA': '41.0.0.0/11',
4860 'ZM': '102.144.0.0/13',
4861 'ZW': '102.177.192.0/18',
4865 def random_ipv4(cls
, code_or_block
):
4866 if len(code_or_block
) == 2:
4867 block
= cls
._country
_ip
_map
.get(code_or_block
.upper())
4871 block
= code_or_block
4872 addr
, preflen
= block
.split('/')
4873 addr_min
= struct
.unpack('!L', socket
.inet_aton(addr
))[0]
4874 addr_max
= addr_min |
(0xffffffff >> int(preflen
))
4875 return str(socket
.inet_ntoa(
4876 struct
.pack('!L', random
.randint(addr_min
, addr_max
))))
4879 class PerRequestProxyHandler(urllib
.request
.ProxyHandler
):
4880 def __init__(self
, proxies
=None):
4881 # Set default handlers
4882 for type in ('http', 'https'):
4883 setattr(self
, '%s_open' % type,
4884 lambda r
, proxy
='__noproxy__', type=type, meth
=self
.proxy_open
:
4885 meth(r
, proxy
, type))
4886 urllib
.request
.ProxyHandler
.__init
__(self
, proxies
)
4888 def proxy_open(self
, req
, proxy
, type):
4889 req_proxy
= req
.headers
.get('Ytdl-request-proxy')
4890 if req_proxy
is not None:
4892 del req
.headers
['Ytdl-request-proxy']
4894 if proxy
== '__noproxy__':
4895 return None # No Proxy
4896 if urllib
.parse
.urlparse(proxy
).scheme
.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4897 req
.add_header('Ytdl-socks-proxy', proxy
)
4898 # yt-dlp's http/https handlers do wrapping the socket with socks
4900 return urllib
.request
.ProxyHandler
.proxy_open(
4901 self
, req
, proxy
, type)
4904 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4905 # released into Public Domain
4906 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4908 def long_to_bytes(n
, blocksize
=0):
4909 """long_to_bytes(n:long, blocksize:int) : string
4910 Convert a long integer to a byte string.
4912 If optional blocksize is given and greater than zero, pad the front of the
4913 byte string with binary zeros so that the length is a multiple of
4916 # after much testing, this algorithm was deemed to be the fastest
4920 s
= struct
.pack('>I', n
& 0xffffffff) + s
4922 # strip off leading zeros
4923 for i
in range(len(s
)):
4924 if s
[i
] != b
'\000'[0]:
4927 # only happens when n == 0
4931 # add back some pad bytes. this could be done more efficiently w.r.t. the
4932 # de-padding being done above, but sigh...
4933 if blocksize
> 0 and len(s
) % blocksize
:
4934 s
= (blocksize
- len(s
) % blocksize
) * b
'\000' + s
4938 def bytes_to_long(s
):
4939 """bytes_to_long(string) : long
4940 Convert a byte string to a long integer.
4942 This is (essentially) the inverse of long_to_bytes().
4947 extra
= (4 - length
% 4)
4948 s
= b
'\000' * extra
+ s
4949 length
= length
+ extra
4950 for i
in range(0, length
, 4):
4951 acc
= (acc
<< 32) + struct
.unpack('>I', s
[i
:i
+ 4])[0]
4955 def ohdave_rsa_encrypt(data
, exponent
, modulus
):
4957 Implement OHDave's RSA algorithm. See http://www.ohdave.com/rsa/
4960 data: data to encrypt, bytes-like object
4961 exponent, modulus: parameter e and N of RSA algorithm, both integer
4962 Output: hex string of encrypted data
4964 Limitation: supports one block encryption only
4967 payload
= int(binascii
.hexlify(data
[::-1]), 16)
4968 encrypted
= pow(payload
, exponent
, modulus
)
4969 return '%x' % encrypted
4972 def pkcs1pad(data
, length
):
4974 Padding input data with PKCS#1 scheme
4976 @param {int[]} data input data
4977 @param {int} length target length
4978 @returns {int[]} padded data
4980 if len(data
) > length
- 11:
4981 raise ValueError('Input data too long for PKCS#1 padding')
4983 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4984 return [0, 2] + pseudo_random
+ [0] + data
4987 def _base_n_table(n
, table
):
4988 if not table
and not n
:
4989 raise ValueError('Either table or n must be specified')
4990 table
= (table
or '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ')[:n
]
4992 if n
and n
!= len(table
):
4993 raise ValueError(f
'base {n} exceeds table length {len(table)}')
4997 def encode_base_n(num
, n
=None, table
=None):
4998 """Convert given int to a base-n string"""
4999 table
= _base_n_table(n
, table
)
5003 result
, base
= '', len(table
)
5005 result
= table
[num
% base
] + result
5010 def decode_base_n(string
, n
=None, table
=None):
5011 """Convert given base-n string to int"""
5012 table
= {char: index for index, char in enumerate(_base_n_table(n, table))}
5013 result
, base
= 0, len(table
)
5015 result
= result
* base
+ table
[char
]
5019 def decode_packed_codes(code
):
5020 mobj
= re
.search(PACKED_CODES_RE
, code
)
5021 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
5024 symbols
= symbols
.split('|')
5029 base_n_count
= encode_base_n(count
, base
)
5030 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
5033 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
5037 def caesar(s
, alphabet
, shift
):
5042 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
5047 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
5050 def parse_m3u8_attributes(attrib
):
5052 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
5053 if val
.startswith('"'):
5059 def urshift(val
, n
):
5060 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
5063 def write_xattr(path
, key
, value
):
5064 # Windows: Write xattrs to NTFS Alternate Data Streams:
5065 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
5066 if compat_os_name
== 'nt':
5067 assert ':' not in key
5068 assert os
.path
.exists(path
)
5071 with open(f
'{path}:{key}', 'wb') as f
:
5073 except OSError as e
:
5074 raise XAttrMetadataError(e
.errno
, e
.strerror
)
5077 # UNIX Method 1. Use xattrs/pyxattrs modules
5080 if getattr(xattr
, '_yt_dlp__identifier', None) == 'pyxattr':
5081 # Unicode arguments are not supported in pyxattr until version 0.5.0
5082 # See https://github.com/ytdl-org/youtube-dl/issues/5498
5083 if version_tuple(xattr
.__version
__) >= (0, 5, 0):
5084 setxattr
= xattr
.set
5086 setxattr
= xattr
.setxattr
5090 setxattr(path
, key
, value
)
5091 except OSError as e
:
5092 raise XAttrMetadataError(e
.errno
, e
.strerror
)
5095 # UNIX Method 2. Use setfattr/xattr executables
5096 exe
= ('setfattr' if check_executable('setfattr', ['--version'])
5097 else 'xattr' if check_executable('xattr', ['-h']) else None)
5099 raise XAttrUnavailableError(
5100 'Couldn\'t find a tool to set the xattrs. Install either the python "xattr" or "pyxattr" modules or the '
5101 + ('"xattr" binary' if sys
.platform
!= 'linux' else 'GNU "attr" package (which contains the "setfattr" tool)'))
5103 value
= value
.decode()
5105 _
, stderr
, returncode
= Popen
.run(
5106 [exe
, '-w', key
, value
, path
] if exe
== 'xattr' else [exe
, '-n', key
, '-v', value
, path
],
5107 text
=True, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
5108 except OSError as e
:
5109 raise XAttrMetadataError(e
.errno
, e
.strerror
)
5111 raise XAttrMetadataError(returncode
, stderr
)
5114 def random_birthday(year_field
, month_field
, day_field
):
5115 start_date
= datetime
.date(1950, 1, 1)
5116 end_date
= datetime
.date(1995, 12, 31)
5117 offset
= random
.randint(0, (end_date
- start_date
).days
)
5118 random_date
= start_date
+ datetime
.timedelta(offset
)
5120 year_field
: str(random_date
.year
),
5121 month_field
: str(random_date
.month
),
5122 day_field
: str(random_date
.day
),
5126 def find_available_port(interface
=''):
5128 with socket
.socket() as sock
:
5129 sock
.bind((interface
, 0))
5130 return sock
.getsockname()[1]
5135 # Templates for internet shortcut files, which are plain text files.
5136 DOT_URL_LINK_TEMPLATE
= '''\
5141 DOT_WEBLOC_LINK_TEMPLATE
= '''\
5142 <?xml version="1.0" encoding="UTF-8"?>
5143 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
5144 <plist version="1.0">
5147 \t<string>%(url)s</string>
5152 DOT_DESKTOP_LINK_TEMPLATE
= '''\
5162 'url': DOT_URL_LINK_TEMPLATE
,
5163 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
5164 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
5168 def iri_to_uri(iri
):
5170 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
5172 The function doesn't add an additional layer of escaping; e.g., it doesn't escape `%3C` as `%253C`. Instead, it percent-escapes characters with an underlying UTF-8 encoding *besides* those already escaped, leaving the URI intact.
5175 iri_parts
= urllib
.parse
.urlparse(iri
)
5177 if '[' in iri_parts
.netloc
:
5178 raise ValueError('IPv6 URIs are not, yet, supported.')
5179 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
5181 # The `safe` argument values, that the following code uses, contain the characters that should not be percent-encoded. Everything else but letters, digits and '_.-' will be percent-encoded with an underlying UTF-8 encoding. Everything already percent-encoded will be left as is.
5184 if iri_parts
.username
:
5185 net_location
+= urllib
.parse
.quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
5186 if iri_parts
.password
is not None:
5187 net_location
+= ':' + urllib
.parse
.quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
5190 net_location
+= iri_parts
.hostname
.encode('idna').decode() # Punycode for Unicode hostnames.
5191 # The 'idna' encoding produces ASCII text.
5192 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
5193 net_location
+= ':' + str(iri_parts
.port
)
5195 return urllib
.parse
.urlunparse(
5199 urllib
.parse
.quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
5201 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
5202 urllib
.parse
.quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
5204 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
5205 urllib
.parse
.quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
5207 urllib
.parse
.quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
5209 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
5212 def to_high_limit_path(path
):
5213 if sys
.platform
in ['win32', 'cygwin']:
5214 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
5215 return '\\\\?\\' + os
.path
.abspath(path
)
5220 def format_field(obj
, field
=None, template
='%s', ignore
=NO_DEFAULT
, default
='', func
=IDENTITY
):
5221 val
= traversal
.traverse_obj(obj
, *variadic(field
))
5222 if not val
if ignore
is NO_DEFAULT
else val
in variadic(ignore
):
5224 return template
% func(val
)
5227 def clean_podcast_url(url
):
5228 return re
.sub(r
'''(?x)
5232 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
5235 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
5238 cn\.co| # https://podcorn.com/analytics-prefix/
5239 st\.fm # https://podsights.com/docs/
5244 _HEX_TABLE
= '0123456789abcdef'
5247 def random_uuidv4():
5248 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
5251 def make_dir(path
, to_screen
=None):
5253 dn
= os
.path
.dirname(path
)
5255 os
.makedirs(dn
, exist_ok
=True)
5257 except OSError as err
:
5258 if callable(to_screen
) is not None:
5259 to_screen(f
'unable to create directory {err}')
5263 def get_executable_path():
5264 from ..update
import _get_variant_and_executable_path
5266 return os
.path
.dirname(os
.path
.abspath(_get_variant_and_executable_path()[1]))
5269 def get_user_config_dirs(package_name
):
5270 # .config (e.g. ~/.config/package_name)
5271 xdg_config_home
= os
.getenv('XDG_CONFIG_HOME') or compat_expanduser('~/.config')
5272 yield os
.path
.join(xdg_config_home
, package_name
)
5274 # appdata (%APPDATA%/package_name)
5275 appdata_dir
= os
.getenv('appdata')
5277 yield os
.path
.join(appdata_dir
, package_name
)
5279 # home (~/.package_name)
5280 yield os
.path
.join(compat_expanduser('~'), f
'.{package_name}')
5283 def get_system_config_dirs(package_name
):
5285 yield os
.path
.join('/etc', package_name
)
5288 def time_seconds(**kwargs
):
5290 Returns TZ-aware time in seconds since the epoch (1970-01-01T00:00:00Z)
5292 return time
.time() + datetime
.timedelta(**kwargs
).total_seconds()
5295 # create a JSON Web Signature (jws) with HS256 algorithm
5296 # the resulting format is in JWS Compact Serialization
5297 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5298 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5299 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5305 header_data
.update(headers
)
5306 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode())
5307 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode())
5308 h
= hmac
.new(key
.encode(), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5309 signature_b64
= base64
.b64encode(h
.digest())
5310 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5314 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5315 def jwt_decode_hs256(jwt
):
5316 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5317 # add trailing ='s that may have been stripped, superfluous ='s are ignored
5318 payload_data
= json
.loads(base64
.urlsafe_b64decode(f
'{payload_b64}==='))
5322 WINDOWS_VT_MODE
= False if compat_os_name
== 'nt' else None
5326 def supports_terminal_sequences(stream
):
5327 if compat_os_name
== 'nt':
5328 if not WINDOWS_VT_MODE
:
5330 elif not os
.getenv('TERM'):
5333 return stream
.isatty()
5334 except BaseException
:
5338 def windows_enable_vt_mode():
5339 """Ref: https://bugs.python.org/issue30075 """
5340 if get_windows_version() < (10, 0, 10586):
5344 import ctypes
.wintypes
5347 ENABLE_VIRTUAL_TERMINAL_PROCESSING
= 0x0004
5349 dll
= ctypes
.WinDLL('kernel32', use_last_error
=False)
5350 handle
= os
.open('CONOUT$', os
.O_RDWR
)
5352 h_out
= ctypes
.wintypes
.HANDLE(msvcrt
.get_osfhandle(handle
))
5353 dw_original_mode
= ctypes
.wintypes
.DWORD()
5354 success
= dll
.GetConsoleMode(h_out
, ctypes
.byref(dw_original_mode
))
5356 raise Exception('GetConsoleMode failed')
5358 success
= dll
.SetConsoleMode(h_out
, ctypes
.wintypes
.DWORD(
5359 dw_original_mode
.value | ENABLE_VIRTUAL_TERMINAL_PROCESSING
))
5361 raise Exception('SetConsoleMode failed')
5365 global WINDOWS_VT_MODE
5366 WINDOWS_VT_MODE
= True
5367 supports_terminal_sequences
.cache_clear()
5370 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5373 def remove_terminal_sequences(string
):
5374 return _terminal_sequences_re
.sub('', string
)
5377 def number_of_digits(number
):
5378 return len('%d' % number
)
5381 def join_nonempty(*values
, delim
='-', from_dict
=None):
5382 if from_dict
is not None:
5383 values
= (traversal
.traverse_obj(from_dict
, variadic(v
)) for v
in values
)
5384 return delim
.join(map(str, filter(None, values
)))
5387 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
5389 Find the largest format dimensions in terms of video width and, for each thumbnail:
5390 * Modify the URL: Match the width with the provided regex and replace with the former width
5393 This function is useful with video services that scale the provided thumbnails on demand
5395 _keys
= ('width', 'height')
5396 max_dimensions
= max(
5397 (tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
),
5399 if not max_dimensions
[0]:
5403 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
5404 dict(zip(_keys
, max_dimensions
)), thumbnail
)
5405 for thumbnail
in thumbnails
5409 def parse_http_range(range):
5410 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
5412 return None, None, None
5413 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
5415 return None, None, None
5416 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
5419 def read_stdin(what
):
5420 eof
= 'Ctrl+Z' if compat_os_name
== 'nt' else 'Ctrl+D'
5421 write_string(f
'Reading {what} from STDIN - EOF ({eof}) to end:\n')
5425 def determine_file_encoding(data
):
5427 Detect the text encoding used
5428 @returns (encoding, bytes to skip)
5431 # BOM marks are given priority over declarations
5432 for bom
, enc
in BOMS
:
5433 if data
.startswith(bom
):
5434 return enc
, len(bom
)
5436 # Strip off all null bytes to match even when UTF-16 or UTF-32 is used.
5437 # We ignore the endianness to get a good enough match
5438 data
= data
.replace(b
'\0', b
'')
5439 mobj
= re
.match(rb
'(?m)^#\s*coding\s*:\s*(\S+)\s*$', data
)
5440 return mobj
.group(1).decode() if mobj
else None, 0
5447 __initialized
= False
5449 def __init__(self
, parser
, label
=None):
5450 self
.parser
, self
.label
= parser
, label
5451 self
._loaded
_paths
, self
.configs
= set(), []
5453 def init(self
, args
=None, filename
=None):
5454 assert not self
.__initialized
5455 self
.own_args
, self
.filename
= args
, filename
5456 return self
.load_configs()
5458 def load_configs(self
):
5461 location
= os
.path
.realpath(self
.filename
)
5462 directory
= os
.path
.dirname(location
)
5463 if location
in self
._loaded
_paths
:
5465 self
._loaded
_paths
.add(location
)
5467 self
.__initialized
= True
5468 opts
, _
= self
.parser
.parse_known_args(self
.own_args
)
5469 self
.parsed_args
= self
.own_args
5470 for location
in opts
.config_locations
or []:
5472 if location
in self
._loaded
_paths
:
5474 self
._loaded
_paths
.add(location
)
5475 self
.append_config(shlex
.split(read_stdin('options'), comments
=True), label
='stdin')
5477 location
= os
.path
.join(directory
, expand_path(location
))
5478 if os
.path
.isdir(location
):
5479 location
= os
.path
.join(location
, 'yt-dlp.conf')
5480 if not os
.path
.exists(location
):
5481 self
.parser
.error(f
'config location {location} does not exist')
5482 self
.append_config(self
.read_file(location
), location
)
5486 label
= join_nonempty(
5487 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5489 return join_nonempty(
5490 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5491 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5495 def read_file(filename
, default
=[]):
5497 optionf
= open(filename
, 'rb')
5499 return default
# silently skip if file is not present
5501 enc
, skip
= determine_file_encoding(optionf
.read(512))
5502 optionf
.seek(skip
, io
.SEEK_SET
)
5504 enc
= None # silently skip read errors
5506 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5507 contents
= optionf
.read().decode(enc
or preferredencoding())
5508 res
= shlex
.split(contents
, comments
=True)
5509 except Exception as err
:
5510 raise ValueError(f
'Unable to parse "{filename}": {err}')
5516 def hide_login_info(opts
):
5517 PRIVATE_OPTS
= {'-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'}
5518 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5523 return m
.group('key') + '=PRIVATE'
5527 opts
= list(map(_scrub_eq
, opts
))
5528 for idx
, opt
in enumerate(opts
):
5529 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5530 opts
[idx
+ 1] = 'PRIVATE'
5533 def append_config(self
, *args
, label
=None):
5534 config
= type(self
)(self
.parser
, label
)
5535 config
._loaded
_paths
= self
._loaded
_paths
5536 if config
.init(*args
):
5537 self
.configs
.append(config
)
5541 for config
in reversed(self
.configs
):
5542 yield from config
.all_args
5543 yield from self
.parsed_args
or []
5545 def parse_known_args(self
, **kwargs
):
5546 return self
.parser
.parse_known_args(self
.all_args
, **kwargs
)
5548 def parse_args(self
):
5549 return self
.parser
.parse_args(self
.all_args
)
5552 class WebSocketsWrapper
:
5553 """Wraps websockets module to use in non-async scopes"""
5556 def __init__(self
, url
, headers
=None, connect
=True):
5557 self
.loop
= asyncio
.new_event_loop()
5558 # XXX: "loop" is deprecated
5559 self
.conn
= websockets
.connect(
5560 url
, extra_headers
=headers
, ping_interval
=None,
5561 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
5564 atexit
.register(self
.__exit
__, None, None, None)
5566 def __enter__(self
):
5568 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
5571 def send(self
, *args
):
5572 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
5574 def recv(self
, *args
):
5575 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
5577 def __exit__(self
, type, value
, traceback
):
5579 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
5582 self
._cancel
_all
_tasks
(self
.loop
)
5584 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
5585 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
5587 def run_with_loop(main
, loop
):
5588 if not asyncio
.iscoroutine(main
):
5589 raise ValueError(f
'a coroutine was expected, got {main!r}')
5592 return loop
.run_until_complete(main
)
5594 loop
.run_until_complete(loop
.shutdown_asyncgens())
5595 if hasattr(loop
, 'shutdown_default_executor'):
5596 loop
.run_until_complete(loop
.shutdown_default_executor())
5599 def _cancel_all_tasks(loop
):
5600 to_cancel
= asyncio
.all_tasks(loop
)
5605 for task
in to_cancel
:
5608 # XXX: "loop" is removed in python 3.10+
5609 loop
.run_until_complete(
5610 asyncio
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
5612 for task
in to_cancel
:
5613 if task
.cancelled():
5615 if task
.exception() is not None:
5616 loop
.call_exception_handler({
5617 'message': 'unhandled exception during asyncio.run() shutdown',
5618 'exception': task
.exception(),
5623 def merge_headers(*dicts
):
5624 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
5625 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
5628 def cached_method(f
):
5629 """Cache a method"""
5630 signature
= inspect
.signature(f
)
5633 def wrapper(self
, *args
, **kwargs
):
5634 bound_args
= signature
.bind(self
, *args
, **kwargs
)
5635 bound_args
.apply_defaults()
5636 key
= tuple(bound_args
.arguments
.values())[1:]
5638 cache
= vars(self
).setdefault('_cached_method__cache', {}).setdefault(f.__name__, {}
)
5639 if key
not in cache
:
5640 cache
[key
] = f(self
, *args
, **kwargs
)
5645 class classproperty
:
5646 """property access for class methods with optional caching"""
5647 def __new__(cls
, func
=None, *args
, **kwargs
):
5649 return functools
.partial(cls
, *args
, **kwargs
)
5650 return super().__new
__(cls
)
5652 def __init__(self
, func
, *, cache
=False):
5653 functools
.update_wrapper(self
, func
)
5655 self
._cache
= {} if cache
else None
5657 def __get__(self
, _
, cls
):
5658 if self
._cache
is None:
5659 return self
.func(cls
)
5660 elif cls
not in self
._cache
:
5661 self
._cache
[cls
] = self
.func(cls
)
5662 return self
._cache
[cls
]
5665 class function_with_repr
:
5666 def __init__(self
, func
, repr_
=None):
5667 functools
.update_wrapper(self
, func
)
5668 self
.func
, self
.__repr
= func
, repr_
5670 def __call__(self
, *args
, **kwargs
):
5671 return self
.func(*args
, **kwargs
)
5676 return f
'{self.func.__module__}.{self.func.__qualname__}'
5679 class Namespace(types
.SimpleNamespace
):
5680 """Immutable namespace"""
5683 return iter(self
.__dict
__.values())
5687 return self
.__dict
__.items()
5690 MEDIA_EXTENSIONS
= Namespace(
5691 common_video
=('avi', 'flv', 'mkv', 'mov', 'mp4', 'webm'),
5692 video
=('3g2', '3gp', 'f4v', 'mk3d', 'divx', 'mpg', 'ogv', 'm4v', 'wmv'),
5693 common_audio
=('aiff', 'alac', 'flac', 'm4a', 'mka', 'mp3', 'ogg', 'opus', 'wav'),
5694 audio
=('aac', 'ape', 'asf', 'f4a', 'f4b', 'm4b', 'm4p', 'm4r', 'oga', 'ogx', 'spx', 'vorbis', 'wma', 'weba'),
5695 thumbnails
=('jpg', 'png', 'webp'),
5696 storyboards
=('mhtml', ),
5697 subtitles
=('srt', 'vtt', 'ass', 'lrc'),
5698 manifests
=('f4f', 'f4m', 'm3u8', 'smil', 'mpd'),
5700 MEDIA_EXTENSIONS
.video
+= MEDIA_EXTENSIONS
.common_video
5701 MEDIA_EXTENSIONS
.audio
+= MEDIA_EXTENSIONS
.common_audio
5703 KNOWN_EXTENSIONS
= (*MEDIA_EXTENSIONS
.video
, *MEDIA_EXTENSIONS
.audio
, *MEDIA_EXTENSIONS
.manifests
)
5708 for retry in RetryManager(...):
5711 except SomeException as err:
5715 attempt
, _error
= 0, None
5717 def __init__(self
, _retries
, _error_callback
, **kwargs
):
5718 self
.retries
= _retries
or 0
5719 self
.error_callback
= functools
.partial(_error_callback
, **kwargs
)
5721 def _should_retry(self
):
5722 return self
._error
is not NO_DEFAULT
and self
.attempt
<= self
.retries
5726 if self
._error
is NO_DEFAULT
:
5731 def error(self
, value
):
5735 while self
._should
_retry
():
5736 self
.error
= NO_DEFAULT
5740 self
.error_callback(self
.error
, self
.attempt
, self
.retries
)
5743 def report_retry(e
, count
, retries
, *, sleep_func
, info
, warn
, error
=None, suffix
=None):
5744 """Utility function for reporting retries"""
5747 return error(f
'{e}. Giving up after {count - 1} retries') if count
> 1 else error(str(e
))
5752 elif isinstance(e
, ExtractorError
):
5753 e
= remove_end(str_or_none(e
.cause
) or e
.orig_msg
, '.')
5754 warn(f
'{e}. Retrying{format_field(suffix, None, " %s")} ({count}/{retries})...')
5756 delay
= float_or_none(sleep_func(n
=count
- 1)) if callable(sleep_func
) else sleep_func
5758 info(f
'Sleeping {delay:.2f} seconds ...')
5762 def make_archive_id(ie
, video_id
):
5763 ie_key
= ie
if isinstance(ie
, str) else ie
.ie_key()
5764 return f
'{ie_key.lower()} {video_id}'
5767 def truncate_string(s
, left
, right
=0):
5768 assert left
> 3 and right
>= 0
5769 if s
is None or len(s
) <= left
+ right
:
5771 return f
'{s[:left-3]}...{s[-right:] if right else ""}'
5774 def orderedSet_from_options(options
, alias_dict
, *, use_regex
=False, start
=None):
5775 assert 'all' in alias_dict
, '"all" alias is required'
5776 requested
= list(start
or [])
5778 discard
= val
.startswith('-')
5782 if val
in alias_dict
:
5783 val
= alias_dict
[val
] if not discard
else [
5784 i
[1:] if i
.startswith('-') else f
'-{i}' for i
in alias_dict
[val
]]
5785 # NB: Do not allow regex in aliases for performance
5786 requested
= orderedSet_from_options(val
, alias_dict
, start
=requested
)
5789 current
= (filter(re
.compile(val
, re
.I
).fullmatch
, alias_dict
['all']) if use_regex
5790 else [val
] if val
in alias_dict
['all'] else None)
5792 raise ValueError(val
)
5795 for item
in current
:
5796 while item
in requested
:
5797 requested
.remove(item
)
5799 requested
.extend(current
)
5801 return orderedSet(requested
)
5805 regex
= r
' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
5807 default
= ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
5808 'res', 'fps', 'hdr:12', 'vcodec:vp9.2', 'channels', 'acodec',
5809 'size', 'br', 'asr', 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
5810 ytdl_default
= ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
5811 'height', 'width', 'proto', 'vext', 'abr', 'aext',
5812 'fps', 'fs_approx', 'source', 'id')
5815 'vcodec': {'type': 'ordered', 'regex': True,
5816 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
5817 'acodec': {'type': 'ordered', 'regex': True,
5818 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'ac-?4', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
5819 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
5820 'order': ['dv', '(hdr)?12', r
'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
5821 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
5822 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
5823 'vext': {'type': 'ordered', 'field': 'video_ext',
5824 'order': ('mp4', 'mov', 'webm', 'flv', '', 'none'),
5825 'order_free': ('webm', 'mp4', 'mov', 'flv', '', 'none')},
5826 'aext': {'type': 'ordered', 'regex': True, 'field': 'audio_ext',
5827 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'web[am]', '', 'none'),
5828 'order_free': ('ogg', 'opus', 'web[am]', 'mp3', 'm4a', 'aac', '', 'none')},
5829 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000}
,
5830 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
5831 'field': ('vcodec', 'acodec'),
5832 'function': lambda it
: int(any(v
!= 'none' for v
in it
))},
5833 'ie_pref': {'priority': True, 'type': 'extractor'}
,
5834 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5835 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5836 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1}
,
5837 'quality': {'convert': 'float', 'default': -1}
,
5838 'filesize': {'convert': 'bytes'}
,
5839 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'}
,
5840 'id': {'convert': 'string', 'field': 'format_id'}
,
5841 'height': {'convert': 'float_none'}
,
5842 'width': {'convert': 'float_none'}
,
5843 'fps': {'convert': 'float_none'}
,
5844 'channels': {'convert': 'float_none', 'field': 'audio_channels'}
,
5845 'tbr': {'convert': 'float_none'}
,
5846 'vbr': {'convert': 'float_none'}
,
5847 'abr': {'convert': 'float_none'}
,
5848 'asr': {'convert': 'float_none'}
,
5849 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1}
,
5851 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')}
,
5852 'br': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True}
,
5853 'size': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')}
,
5854 'ext': {'type': 'combined', 'field': ('vext', 'aext')}
,
5855 'res': {'type': 'multiple', 'field': ('height', 'width'),
5856 'function': lambda it
: (lambda l
: min(l
) if l
else 0)(tuple(filter(None, it
)))},
5858 # Actual field names
5859 'format_id': {'type': 'alias', 'field': 'id'}
,
5860 'preference': {'type': 'alias', 'field': 'ie_pref'}
,
5861 'language_preference': {'type': 'alias', 'field': 'lang'}
,
5862 'source_preference': {'type': 'alias', 'field': 'source'}
,
5863 'protocol': {'type': 'alias', 'field': 'proto'}
,
5864 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'}
,
5865 'audio_channels': {'type': 'alias', 'field': 'channels'}
,
5868 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5869 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5870 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True}
,
5871 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True}
,
5872 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True}
,
5873 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True}
,
5874 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True}
,
5875 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True}
,
5876 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True}
,
5877 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True}
,
5878 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True}
,
5879 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True}
,
5880 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True}
,
5881 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True}
,
5882 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5883 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5884 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5885 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5886 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5887 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5890 def __init__(self
, ydl
, field_preference
):
5893 self
.evaluate_params(self
.ydl
.params
, field_preference
)
5894 if ydl
.params
.get('verbose'):
5895 self
.print_verbose_info(self
.ydl
.write_debug
)
5897 def _get_field_setting(self
, field
, key
):
5898 if field
not in self
.settings
:
5899 if key
in ('forced', 'priority'):
5901 self
.ydl
.deprecated_feature(f
'Using arbitrary fields ({field}) for format sorting is '
5902 'deprecated and may be removed in a future version')
5903 self
.settings
[field
] = {}
5904 propObj
= self
.settings
[field
]
5905 if key
not in propObj
:
5906 type = propObj
.get('type')
5908 default
= 'preference' if type == 'extractor' else (field
,) if type in ('combined', 'multiple') else field
5909 elif key
== 'convert':
5910 default
= 'order' if type == 'ordered' else 'float_string' if field
else 'ignore'
5912 default
= {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}
.get(key
, None)
5913 propObj
[key
] = default
5916 def _resolve_field_value(self
, field
, value
, convertNone
=False):
5921 value
= value
.lower()
5922 conversion
= self
._get
_field
_setting
(field
, 'convert')
5923 if conversion
== 'ignore':
5925 if conversion
== 'string':
5927 elif conversion
== 'float_none':
5928 return float_or_none(value
)
5929 elif conversion
== 'bytes':
5930 return parse_bytes(value
)
5931 elif conversion
== 'order':
5932 order_list
= (self
._use
_free
_order
and self
._get
_field
_setting
(field
, 'order_free')) or self
._get
_field
_setting
(field
, 'order')
5933 use_regex
= self
._get
_field
_setting
(field
, 'regex')
5934 list_length
= len(order_list
)
5935 empty_pos
= order_list
.index('') if '' in order_list
else list_length
+ 1
5936 if use_regex
and value
is not None:
5937 for i
, regex
in enumerate(order_list
):
5938 if regex
and re
.match(regex
, value
):
5939 return list_length
- i
5940 return list_length
- empty_pos
# not in list
5941 else: # not regex or value = None
5942 return list_length
- (order_list
.index(value
) if value
in order_list
else empty_pos
)
5944 if value
.isnumeric():
5947 self
.settings
[field
]['convert'] = 'string'
5950 def evaluate_params(self
, params
, sort_extractor
):
5951 self
._use
_free
_order
= params
.get('prefer_free_formats', False)
5952 self
._sort
_user
= params
.get('format_sort', [])
5953 self
._sort
_extractor
= sort_extractor
5955 def add_item(field
, reverse
, closest
, limit_text
):
5956 field
= field
.lower()
5957 if field
in self
._order
:
5959 self
._order
.append(field
)
5960 limit
= self
._resolve
_field
_value
(field
, limit_text
)
5963 'closest': False if limit
is None else closest
,
5964 'limit_text': limit_text
,
5966 if field
in self
.settings
:
5967 self
.settings
[field
].update(data
)
5969 self
.settings
[field
] = data
5972 tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'forced'))
5973 + (tuple() if params
.get('format_sort_force', False)
5974 else tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'priority')))
5975 + tuple(self
._sort
_user
) + tuple(sort_extractor
) + self
.default
)
5977 for item
in sort_list
:
5978 match
= re
.match(self
.regex
, item
)
5980 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item
)
5981 field
= match
.group('field')
5984 if self
._get
_field
_setting
(field
, 'type') == 'alias':
5985 alias
, field
= field
, self
._get
_field
_setting
(field
, 'field')
5986 if self
._get
_field
_setting
(alias
, 'deprecated'):
5987 self
.ydl
.deprecated_feature(f
'Format sorting alias {alias} is deprecated and may '
5988 f
'be removed in a future version. Please use {field} instead')
5989 reverse
= match
.group('reverse') is not None
5990 closest
= match
.group('separator') == '~'
5991 limit_text
= match
.group('limit')
5993 has_limit
= limit_text
is not None
5994 has_multiple_fields
= self
._get
_field
_setting
(field
, 'type') == 'combined'
5995 has_multiple_limits
= has_limit
and has_multiple_fields
and not self
._get
_field
_setting
(field
, 'same_limit')
5997 fields
= self
._get
_field
_setting
(field
, 'field') if has_multiple_fields
else (field
,)
5998 limits
= limit_text
.split(':') if has_multiple_limits
else (limit_text
,) if has_limit
else tuple()
5999 limit_count
= len(limits
)
6000 for (i
, f
) in enumerate(fields
):
6001 add_item(f
, reverse
, closest
,
6002 limits
[i
] if i
< limit_count
6003 else limits
[0] if has_limit
and not has_multiple_limits
6006 def print_verbose_info(self
, write_debug
):
6008 write_debug('Sort order given by user: %s' % ', '.join(self
._sort
_user
))
6009 if self
._sort
_extractor
:
6010 write_debug('Sort order given by extractor: %s' % ', '.join(self
._sort
_extractor
))
6011 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
6012 '+' if self
._get
_field
_setting
(field
, 'reverse') else '', field
,
6013 '%s%s(%s)' % ('~' if self
._get
_field
_setting
(field
, 'closest') else ':',
6014 self
._get
_field
_setting
(field
, 'limit_text'),
6015 self
._get
_field
_setting
(field
, 'limit'))
6016 if self
._get
_field
_setting
(field
, 'limit_text') is not None else '')
6017 for field
in self
._order
if self
._get
_field
_setting
(field
, 'visible')]))
6019 def _calculate_field_preference_from_value(self
, format
, field
, type, value
):
6020 reverse
= self
._get
_field
_setting
(field
, 'reverse')
6021 closest
= self
._get
_field
_setting
(field
, 'closest')
6022 limit
= self
._get
_field
_setting
(field
, 'limit')
6024 if type == 'extractor':
6025 maximum
= self
._get
_field
_setting
(field
, 'max')
6026 if value
is None or (maximum
is not None and value
>= maximum
):
6028 elif type == 'boolean':
6029 in_list
= self
._get
_field
_setting
(field
, 'in_list')
6030 not_in_list
= self
._get
_field
_setting
(field
, 'not_in_list')
6031 value
= 0 if ((in_list
is None or value
in in_list
) and (not_in_list
is None or value
not in not_in_list
)) else -1
6032 elif type == 'ordered':
6033 value
= self
._resolve
_field
_value
(field
, value
, True)
6035 # try to convert to number
6036 val_num
= float_or_none(value
, default
=self
._get
_field
_setting
(field
, 'default'))
6037 is_num
= self
._get
_field
_setting
(field
, 'convert') != 'string' and val_num
is not None
6041 return ((-10, 0) if value
is None
6042 else (1, value
, 0) if not is_num
# if a field has mixed strings and numbers, strings are sorted higher
6043 else (0, -abs(value
- limit
), value
- limit
if reverse
else limit
- value
) if closest
6044 else (0, value
, 0) if not reverse
and (limit
is None or value
<= limit
)
6045 else (0, -value
, 0) if limit
is None or (reverse
and value
== limit
) or value
> limit
6046 else (-1, value
, 0))
6048 def _calculate_field_preference(self
, format
, field
):
6049 type = self
._get
_field
_setting
(field
, 'type') # extractor, boolean, ordered, field, multiple
6050 get_value
= lambda f
: format
.get(self
._get
_field
_setting
(f
, 'field'))
6051 if type == 'multiple':
6052 type = 'field' # Only 'field' is allowed in multiple for now
6053 actual_fields
= self
._get
_field
_setting
(field
, 'field')
6055 value
= self
._get
_field
_setting
(field
, 'function')(get_value(f
) for f
in actual_fields
)
6057 value
= get_value(field
)
6058 return self
._calculate
_field
_preference
_from
_value
(format
, field
, type, value
)
6060 def calculate_preference(self
, format
):
6061 # Determine missing protocol
6062 if not format
.get('protocol'):
6063 format
['protocol'] = determine_protocol(format
)
6065 # Determine missing ext
6066 if not format
.get('ext') and 'url' in format
:
6067 format
['ext'] = determine_ext(format
['url'])
6068 if format
.get('vcodec') == 'none':
6069 format
['audio_ext'] = format
['ext'] if format
.get('acodec') != 'none' else 'none'
6070 format
['video_ext'] = 'none'
6072 format
['video_ext'] = format
['ext']
6073 format
['audio_ext'] = 'none'
6074 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
6075 # format['preference'] = -1000
6077 if format
.get('preference') is None and format
.get('ext') == 'flv' and re
.match('[hx]265|he?vc?', format
.get('vcodec') or ''):
6078 # HEVC-over-FLV is out-of-spec by FLV's original spec
6079 # ref. https://trac.ffmpeg.org/ticket/6389
6080 # ref. https://github.com/yt-dlp/yt-dlp/pull/5821
6081 format
['preference'] = -100
6083 # Determine missing bitrates
6084 if format
.get('tbr') is None:
6085 if format
.get('vbr') is not None and format
.get('abr') is not None:
6086 format
['tbr'] = format
.get('vbr', 0) + format
.get('abr', 0)
6088 if format
.get('vcodec') != 'none' and format
.get('vbr') is None:
6089 format
['vbr'] = format
.get('tbr') - format
.get('abr', 0)
6090 if format
.get('acodec') != 'none' and format
.get('abr') is None:
6091 format
['abr'] = format
.get('tbr') - format
.get('vbr', 0)
6093 return tuple(self
._calculate
_field
_preference
(format
, field
) for field
in self
._order
)